November 2016

INDIAN RAILWAYS INSTITUTE OF CIVIL ENGINEERING

PUNE 411001
Published By
Indian Railways Institute of Civil Engg.
11-A, South Main Road, Koregaon Park, Pune 411 001.

FIRST EDITION - MARCH 1983

SECOND EDITION - MARCH 1998
THIRD EDITION - JUNE 2007
FOURTH EDITION - NOVEMBER 2016

Price 90/-

Printed by
Kalyani Corporation
Pune
 RIVER TRAINING AND
PROTECTION WORKS
FOR RAILWAY BRIDGES

November 2016
INDIAN RAILWAYS INSTITUTE OF CIVIL ENGINEERING
PUNE 411 001
(ii)
 FOREWORD TO FOURTH EDITION

The book “River Training and Protection Works for

Railway Bridges” was first published in year 1983.The
revised fourth edition has been brought making some
changes taking into account the current sub structure code
and Indian Railway bridge manual in chapter 5 and 6.

It is hoped that the book will be found useful by the

field Engineers, involved in the planning, design &
construction of railway bridges.

The suggestions for improvement are welcome.

November 2016 N. C. Sharda

Director/IRICEN
director@iricen.gov.in

(iii)
(iv)
 PREFACE TO FOURTH EDITION

The book “River Training and Protection Works for

Railway Bridges” was first published in year 1983.
Subsequently, it was revised and enlarged as 2nd edition in
1998 and 3rd edition in 2007.

In this 4th edition following changes have been made -

In chapter 5, the instructions regarding guide bunds,

spurs/groyens, approach banks, marginal bunds and
closure bunds have been included as contained in IRBM. &
in chapter 6 the guidelines regarding estimation of Design
Discharge for bridges as contained in Sub-structure Code
have been included.

It is expected that it would be useful for railway

engineers.

Above all, I am grateful to Shri N. C. Sharda, Director

IRICEN for his encouragement and guidance in bringing out
the publication.

November 2016 (Ramesh Pinjani)

Sr. Professor-Bridge 2
IRICEN

(v)
 PREFACE TO THE THIRD REVISED EDITION

The second edition of the book “River Training

and Protection Works for Railway Bridges” was brought
out nine years ago. Keeping in view the popularity of the
book among field engineers and to keep them abreast of
the latest developments taking place in this field, the third
revised and enlarged edition has been brought out by
adding new chapters on Remote Sensing, Filters,
Permeable Structures and Mathematical Modelling so as to
provide latest techniques being used in design of river
training and protection works for bridges.

Although every effort has been made to present the

book in error free manner, yet if there is any suggestion or
discrepancies, kindly do write to us.

Shiv Kumar
Director
IRICEN

(vi)
 ACKNOWLEDGEMENT TO THE
THIRD REVISED EDITION

The first edition of the book on “River Training and

Protection Works for Railway Bridges” was published in
1983. It was written by Dr. S. V. Chitale, retired joint
Director of Central Water and Power Research Station,
Khadakwasla, Pune. It was extremely popular amongst field
engineers. The second revised edition of this book was
brought out in 1998 by making some changes necessitated
due to changes in Substructure and Foundation code.

A need was being felt to revise the book keeping in

view of rapid developments taking place in the field of River
Training and Protection Works, especially in the areas
where latest technology, viz. remote sensing, geo-fabric
filters can be used extensively and effectively for designing
river training and protection works. Age old technique of
channelling flow of rivers with the help of permeable
structures has been added as a separate chapter. In this
revised & enlarged edition, a new chapter on Mathematical
Modelling has been added. In addition, a number of actual
colour photographs have been incorporated in the book for
better understanding of the subject.

The revised edition has been brought out with a

view to provide maximum information to users. In this effort,
Shri R. A. Oak, retired Chief Research Officer, Central
Water & Power Research Station, Khadakwasla, Pune has
provided invaluable assistance by penning new chapters, I
thank him for his valuable contribution. I also thank Shri
Ganesh Srinivasan for providing assistance in word
processing.

Above all, I am grateful to Shri Shiv Kumar,

Director, IRICEN for his encouragement and valuable
guidance in bringing out this publication.
Ajay Goyal
Senior Professor/Bridges
IRICEN

(vii)
 PREFACE TO THE SECOND REVISED EDITION

River training and protection works for railway

bridges play a major role during planning, construction and
service of these bridges. The design of various components
of training and protection works was developed over a pe-
riod of about 100 years and the present methods are time
tested. The first edition of this book was published in 1983
and was very popular amongst Railway Engineers.

This revised second edition has been brought out

making some changes taking into account the current Sub-
structures and Foundations Code. The utility of the book
has been enhanced by improving the presentation of text
matter and sketches. I hope this book will be very useful to
Railway Engineers in design, construction and maintenance
of river training and protection works for railway bridges.

March 1998 Director

Indian Railways Institute of
Civil Engineering,
Pune

(viii)
 PREFACE TO THE FIRST EDITION

Hydraulic design of a bridge comprises several

aspects such as design discharge, waterway, constriction
by approaches, guide bunds, protection works, etc.
Literature available on these topics is mostly scattered and
few publications have attempted to bring important relevant
information at one place. The classic work of Spring (1903)
‘River Training and Control on Guide Bank System’ and
another noteworthy contribution made by Gales (1938)
provide the designer with authoritative guide but substantial
new information has become available subsequently. ‘River
Training and Control of Bridges’ by Sethi (1960) is more
recent and brings together views of Spring; Gales and
others along with his own but still warrants updating.
‘Behavior and control of Rivers and Canals’ by Inglis (1956
- Revised 1971) and ‘Manual on River Behavior, Control
and Training’ are two more outstanding treatises of great
importance but encompass a very wide range of allied
subjects and a bridge engineer has to wade through the
entire work to dig out the required bit of information. A
cogent and coordinated treatment of all topics of interest to
a bridge engineer in the light of present day knowledge on
the subject is long felt wanting. Moreover many of the
publications cited before are now out of print and stock.
The present work strives to fill in this lacuna.

The subject matter is presented in a logical

sequence starting with basic background information about
types of rivers and their characteristics contained in the first
two chapters. Some essential topics like velocity
distribution, flow formulae, afflux, scour, parametric
relationship evolved by Lacey, are dealt with in the next
chapter under ‘Principles of River Training’. The next
chapter entitled “Types of Training and Protection Works”
covers subheads concerning guide bunds, bridge
approaches, their protection by spurs and revetment and
pier protection aspects in general. Concise design norms
for these very components are presented in the subsequent
chapter. The last two chapters touch upon the subjects of

(ix)
hydraulic modeling and measurement of river water levels
and discharges. The entire subject matter is thus grouped
under the above seven chapters.

Dr. S. V. Chitale, retired Joint Director of Central

Water and Power Research Station, Khadakwasla, Pune
has written this monogram, which is being published by the
Indian Railways Institute of Advanced Track Technology,
Pune. It is hoped that this will be useful to all Railway
Engineers and other Engineers who have to design,
construct, inspect and maintain River Training and
Protection works for bridges.

Principal
Indian Railways Institute of
Advanced Track Technology, Pune

(x)
 CONTENTS

1. TYPE OF RIVERS 1
1.1 Upper reaches 1
1.2 Submontane reaches 4
1.3 Flood plain reaches 5
1.3.1 Stable type 5
1.3.2 Aggrading type 6
1.3.3 Degrading type 6
1.4 Tidal reaches 6
1.5 Other types 7
1.5.1 Flashy Rivers 7
1.5.2 Virgin Rivers 8
1.5.3 Himalayan Rivers 9
1.5.4 South Indian Rivers 9

2. BEHAVIOR OF RIVERS 10
2.1 Meandering Pattern 11
2.1.1 Meander Characteristics and
causative factors 12
2.1.2 Cut offs and movement of meanders 13
2.1.3 Meander relationships 16
2.1.4 Engineering implications of meanders 20
2.2 Braiding Pattern 22
2.3 Bed Forms 24
2.4 River Formulae 30

3. REMOTE SENSING AND ITS APPLICATION 37

TO BRIDGE ENGINEERING
3.1 Remote sensing technique 38
3.1.1 Satellites 38
3.1.2 Scanning process 38
3.2 Advantages 40
3.3 Characteristics of data 42

(xi)
 3.4 Acquisition and processing of data 44
3.5 Study of river morphology 45
3.5.1 Interpretation and analysis using
remote sensing 45
3.6 Procurement of data 47
3.6.1 Steps for the procurement 47
3.6.2 Details for the requisition 48
3.7 Concluding Remarks 50

4. PRINCIPLES OF RIVER TRAINING AND

PROTECTION WORKS 52
4.1 Velocity Distribution and Flow Formulae 53
4.2 Afflux in Rivers with Non scourable Bed 55
4.2.1 Afflux due to constriction of waterway 55
4.2.2 Afflux on account of obstruction to flow 58
4.3 Afflux and Scour in Alluvial Rivers 58
4.4 Free Board 59
4.5 Clearance 60
4.6 Lacey Formulae 61

5. TYPES OF TRAINING AND PROTECTION

WORKS 66
5.1 Bridge location and alignment 66
5.2 Bridge waterway 67
5.3 Guide Bunds 69
5.3.1 Shape of guide bunds 69
5.3.2 Length of guide bunds 71
5.3.3 Curved heads 74
5.3.4 Protection for side slopes 79
5.3.5 Apron Protection 90
5.3.5.1 Scour Depth 90
5.3.5.2 Scour Factors 90
5.3.5.3 Apron Quantity 93
5.3.5.4 Apron as laid 95
5.3.5.5 Construction programme 95

(xii)
 5.3.6 Provisions as contained in IRBM 95
para 810 with regards to guide bunds
5.3.6.1 Necessity 95
5.3.6.2 Shape and design features 96
5.3.6.3 Apron protection for guide 98
bunds
5.3.6.4 Maintenance 99
5.3.6.5 Failure and remedial measures 99
5.4 Spurs on Groynes 100
5.4.1 Functions 100
5.4.2 Types 100
5.4.3 Location 104
5.4.4 Length, Spacing, Inclination and
Height 105
5.4.5 Materials of construction 106
5.4.6 Provisions as contained in IRBM 107
para 811 with regards to spurs/groynes
5.4.6.1 Necessity 107
5.4.6.2 Types of spurs/groynes 107
5.4.6.3 Location and salient features 108
of a spur/groynes
5.4.6.4 Maintenance of spurs/groynes 109
5.4.6.5 Permeable structures 109
5.5 Approach banks and Marginal embankments 112
5.5.1 Approach Banks 112
5.5.1.1 Provisions as contained in 113
IRBM para 817 with regards to
protection of approach banks
5.5.2 Marginal embankments 114
5.5.3 Provisions as contained in IRBM
para 812 with regards to marginal
bunds 115
5.6 Bank revetment 115
5.7 Closure bunds 117

(xiii)
 5.7.1 Provisions as contained in IRBM 117
para 813 with regards to closure
bunds
5.8 Artificial cut offs 117
5.8.1 Provisions as contained in IRBM 118
para 814 with regards to assisted/
artificial cut-offs
5.9 Protection for shallow piers 119

5.10 Training and Protection works in hilly and

Submontane rivers 119

6. DESIGN OF TRAINING AND PROTECTION

WORKS 122
6.1 Design Discharge 122
6.1.1 Estimation of design discharge 122
6.1.1.1 New lines and rebuilding 122
of bridges
6.1.1.2 Doubling works and gauge 123
conversion projects
6.1.2 Design discharge for foundation 124
6.2 Design of waterway 125
6.3 Guide bunds 127
6.3.1 Form in Plan 127
6.3.2 Length of guide bund 127
6.3.3 Curved heads 129
6.3.4 Cross section and protection 129
(i) Top width 129
(ii) Top level 129
(iii) Side slopes 130
(iv) Material 130
(v) Size of pitching stone 130
(vi) Thickness of pitching 132
6.3.5 Launching Apron 133

(xiv)
 (i) Size of stone 133
(ii) Thickness 133
(iii) Slope of launched face 133
(iv) Quantity of apron stone 134
(v) Laying of apron 135
6.4 Spurs or Groynes 135
6.4.1 Types of spurs 135
6.4.2 Location 135
6.4.3 Length, inclination, height and
spacing for impermeable spurs 137
6.4.4 Cross section and Protection of
impermeable spurs 137
6.4.5 Permeable spurs 141
6.5 Approach Bank and Marginal Embankments 141
6.5.1 Approach banks 141
6.5.2 Marginal embankments 144
6.6 Bank Revetments 145
6.6.1 Side slope 145
6.6.2 Velocity in the eroding bend along
concave bank 145
6.6.3 Side slope and apron protection 146
6.7 Artificial cut offs 146
6.8 Bridge pier Protection 147
6.8.1 Deep piers 147
(i) Scour depth and grip length 147
(ii) Flood scour 153
(iii) Scour due to constriction 154
(iv) Effect of bed forms 154
(v) Pier scour in clayey strata 154
6.8.2 Shallow piers 156
6.9 Design for protection works: instructions as 157
contained in IRBM para 818
6.10 Data Requirement 158
6.10.1Survey data 158

(xv)
 6.10.2Hydraulic data 159
6.10.3Sediment data 159

7. PLANNING AND DESIGNING OF PERMEABLE

STRUCTURES 162
7.1 Function 163
7.2 Classification 163
7.3 Permeable structures 163
7.3.1 Elements 164
a. Porcupines 164
b. Cribs 165
c. Bally frames 167
d. Tree branches 167
7.3.2 Construction material 168
7.4 Design concepts 168
7.4.1 Selection of elements 169
7.4.2 Layout in plan 169
7.5 Stability of structures 171
7.6 Protection to the structures 172
7.7 Miscellaneous 172

8. FILTERS BELOW STONE PROTECTION

WORKS 174
8.1 Concept of filter design 174
8.2 Filter below the revetment 175
8.3 Design of filters 175
8.3.1 Graded filter 175
8.3.2 Geotextile filter 176
8.4 Design concept 177
8.4.1 Properties 178
8.4.2 Design criteria 179
8.5 Method of construction 179

(xvi)
9. HYDRAULIC MODELLING 182
9.1 Modeling Technique 182
9.2 Similitude Requirement 182
9.3 Model Scale Design 183
9.3.1 Rigid Bed Models 183
9.3.2 Mobile Bed Models 184
9.4 Model Limitations 185
9.5 Dimensional Analysis 185
9.6 Requirement of field data 186
9.6.1 General 186
9.6.2 Survey Data 186
9.6.3 Hydrographic data 187
9.6.4 Sediment Data 187

10. MEASUREMENT OF RIVER WATER LEVELS

AND DISCHARGES 190
10.1 Water level observations 190
10.1.1 Site Requirements 190
10.1.2 Different types of Gauges 190
10.1.3 Crest gauges 192
10.2 Discharge measurements 192
10.2.1 Velocity-Area Method 192
10.2.2 Slope-Area Method 202
10.2.3 Stage Discharge relationship 204
10.2.4 Other Methods 207

11. MATHEMATICAL MODELING 212

11.1 Type of models 213
11.2 Systematic stages for the studies 218
11.3 Advantages and disadvantages 218
a. Scale effects 219
b. Type of models utilised 219
c. Cost and time 219
d. Establishment cost 219

(xvii)
 e. Model limitations 220
11.4 Closing remarks 220

List of Figures & Photos 222

List of Tables 226
Index 228

(xviii)
 Chapter 1

TYPES OF RIVERS
In Bridge Engineering, intimate knowledge of the
environmental medium is as important as that of the
structure since their interaction governs safety of the bridges
and final regime of the river.

For studying the characteristics and behaviour of

rivers, some classification is required to be followed. Basis
for such classification could be varied, as for instance,
according to the -

- Nature of bed and bank material

- River flood hydrograph,
- Geological age of river,
- Overall river characteristics, etc.

Out of these different classification schemes, the last

one is preferred, being the most convenient and appropriate
in the context of the subject of bridge hydraulics. In the
same river, characteristics go on changing from the head
reach in the hills towards the tail end at the outfall into the
sea. Suitable division of a river system can, therefore, be
made as below :-

(a) Upper (upland hilly) reaches

(b) Submontane (foothill) reaches,
(c) Flood plain reaches,
(d) Tidal reaches, and
(e) Other types.

These are discussed below :

1.1 UPPER REACHES

In the hilly upland reach, the river course and its plan
form are mostly governed by hill contours. Rocky banks
impose side constraint in lateral movement of the river. The
riverbed comprises boulders, shingle, gravel and material of

1
Photo 1.1 : River Ravi (left) and River Pubbar (right) in
the hilly region of Himachal Pradesh. Note : Banks of
the river are fully governed by the hills. The bed
material is comprised of boulders of various sizes.

smaller sizes. Such heterogeneous material cannot develop

bed form of ripples, dunes or waves. River gorge is often
deep and narrow and river slope very steep resulting in
formation of frequent rapids. River velocity is high and
capable of transporting big boulders, which can cause rapid
abrasion and damage to structural components such as
bridge piers due to friction and impact. Photo 1.1 shows
River Ravi and River Pubber in the hilly regions of Himachal
Pradesh.

Floods in hilly terrain are generally flashy. Because

of landslides, damming up of river flow can occur at times.
When such a dam bursts because of overtopping, a sudden
flood wave of catastrophic magnitude can sweep down the
river. Sutlej River in Himachal Pradesh experienced such
catastrophic floods in August / September 2000, due to dam
burst in Tibet. It was reported that about 60 bridges located
on the river were lost in that catastrophe. Photo 1.2 shows
remnants of a suspension bridge lost during the catastrophe.

2
 Photo 1.2 : Remnants of a suspension bridge lost during
the catastrophe

The sediment transport in upland reach comprises

both heavy bed load as well as suspended load. The
velocities being high, sediment transporting capacity and
hence the sediment load is very heavy.

In boulder bedded hilly torrents, the channel may at

times form deposits of bed material during heavy floods due
to soil erosion and land slips. Measures for controlling soil
erosion, improving stability of side slopes and arresting bed
load are found useful under these conditions. Preservation of
soil cover can be promoted by afforestation, gully and check
dams, contour bunding etc. For improving stability of side
slopes, provision of longitudinal and lateral drainage, breast
and toe walls, chutes and in extreme cases bored piles are
found effective. Roads running along upper reaches of some
of the tributaries of the Ganga and the Brahmaputra have
been maintained by providing such elaborate drainage
systems. Excessive bed load, which may lead to choking of
the channel, can be arrested by construction of debris dams
and detention basins. Los Angeles, city in the U.S.A. is
protected from boulders rolling down steep adjoining streams
by means of debris dams. On Gandak River in the
Himalayan gorge, an extensive widening of the valley floor
has provided the river with a natural detention basin where
coarse bed load is arrested.

3
1.2 SUBMONTANE REACHES

The foothill area in between the high mountains and

relatively flat flood plain is termed submontane region. Local
names such as Bhabar area are also in vogue. The river in
submontane reach suddenly acquires much flatter bed
gradient than the hilly mountainous tract. The flow velocity,
therefore, drastically reduces and so its sediment
transporting capacity. This phenomenon encourages
deposition of the excess sediment load. The rivers in the
reach are normally prone to aggradation, which in turn leads
to lateral shifting of the course. An alluvial fan, also termed
as inland delta is built up and progressively it is raised and
extended. Bed and bank material is not much different
except at the extremities of the delta. The bed gradient is
not as steep as in the gorge but not as flat as in the plain.
Relative to lower reaches, the river course being steeper has
got higher velocities. These characteristics develop with
several active channels existing simultaneously intersecting
each other and forming islands in between. All these
channels normally over-flow during high flood and the river
acquire a very wide and shallow cross section. Foothill
reaches of the tributaries of the Ganga and Brahmaputra river

Photo 1.3 : Tributary of River Sutlej in Himachal

Pradesh. Note : The wide range of sediment sizes,
deposited on the riverbed. Existence of old dry channel
along the right and left banks and existing channel in
the middle indicates instability of the river channel.

4
systems are illustrative of braiding pattern. Photo 1.3 shows
a tributary of River Sutlej in Himachal Pradesh. Peak floods
can more often be flashy. Intensity of suspended and bed
load transport is intermediate between upland and flood plain
reaches.

1.3 FLOOD PLAIN REACHES

The river in this reach has a relatively deep and

narrow cross section with medium slope. In high floods,
extensive heavy spills can occur. These spills result in
deposition on over bank area of suspended sediment having
high fertilising value. Flooding due to such river is thus
beneficial in increasing agricultural productivity. On the other
hand, frequent and prolonged inundation can result in untold
miseries and hardships to the population and sometimes
causing enormous loss of property and even life.

The channel is usually of meandering pattern with

meanders at times moving down-stream or forming cut offs
across the neck. The Khadir width within which all meanders
are contained is several times bigger than the river width at
bankful stage though during high floods, the entire width of
the Khadir is occupied. The river bank material contains a
good percentage of clayey silt and the side slope can
therefore be steep.

The bed is normally sandy and during floods big

sized bed forms can develop like sand dunes and waves,
which move progressively downstream. The floods are
sustained and of relatively long duration as in this reach,
several tributaries may join the parent river and form the river
system.

Sediment concentration is high for fine as well as

medium and coarser sizes. Bed load in comparison with
suspended load is much smaller.

1.3.1 Stable type

Rivers in flood plains are normally stable with no

perceptible rise or lowering of the riverbed occurring from year

5
to year over long periods. Changes in the river bed do occur
from season to season, but on an annual basis, bed level
changes are nominal. Ganga River is a notable example of
such type of rivers.

1.3.2 Aggrading type

On the other hand, some rivers are known to

aggrade and progressively raise their beds by sediment
deposition. In the process, overbank spills increase year after
year until, with occurrence of abnormal floods any year, an
avulsion can take place. The Yellow River in China has been
notorious for such avulsions causing extensive devastation in
its wake. Aggradation can be the result either of increased
sediment yield or reduction in flood discharge or both these
factors.

1.3.3 Degrading type

Opposed to aggrading tendency, the river may be

prone to degrade its bed, if sediment supply reduces or flood
discharge increases. Reduction of sediment supply can be
due to reforestation and land management in the catchment
or because of construction of a reservoir impounding
appreciable quantity of water and the sediment. Colorado
river in the U.S.A downstream of Hoover dam, which was
previously known as Boulder dam, is a well-known example
of degrading river. The river bed scoured by as much as
5.5 m at about 16 km in the first 7 years of completion of
the dam. This lowering reduced progressively in the
downstream direction, the length affected by degradation
being as long as 155 km.

Aggradation reduces available waterway area for

passage of flood flow and causes progressive rise in flood
levels. On the other hand, degradation can affect safety of
foundations. Thus, both aggrading and degrading rivers affect
utility and threaten safety of the structures.

1.4 TIDAL REACHES

At the outfall end of the river system, the effect of

6
sea tides is felt and predominates over the tidal reach. The
tidal effect depends on tidal range, which is the difference
between high water and low water and varies from one tidal
waterway to another. The effect also changes from day to
day due to changing position of the earth with respect to sun
and the moon. On full moon and new moon days, lunar effect
and so the tidal ranges are the highest. These tides are
called spring tides. On the other hand, at quarters lunar
effect is the least and so also the tidal range. These are
called neap tides. When moon is the nearest, the tide is
called perigee tide and when the farthest it is called apogee
tide. As the flood tide progresses from the sea into the
estuary and tidal creeks, shallow water effect is felt and wave
form gets distorted, the rising limb of the tidal wave
becoming steeper. Accordingly, the wave velocity becomes
faster and in an extreme case, the flood tide moves up like a
wall when it is called a bore tide.

Along with changes in water levels, the tides also

cause river waters to become saline. The maximum salinity
is at sea end and progressively reduces towards the
upstream end of a tidal reach. The salinity changes along the
length and according to tidal effect. In addition the fresh
discharge of the river exerts influence on both the tides and
salinity.

Due to salinity and consequent flocculation, fine

suspended sediment deposits in tidal waterways. Deposition
process is also affected by changes in density of water. The
river regime is thus markedly governed by tides and
obstruction caused by a bridge to propagation of tides can
seriously affect the river regime.

1.5 OTHER TYPES

1.5.1 Flashy Rivers

The rivers in upper hilly reaches and submontane

area often experience flashy floods. Heavy storms can cover
and concentrate over relatively smaller catchments of such
rivers. A flood wave of short duration but of high intensity
then rushes down the river like a flash. In such a flood, heavy

7
sediment load can be brought down by the river, which can
be deposited on the way. Flashy floods cannot scour the
riverbed sufficiently at constrictions within guide bunds
thereby causing high afflux. In case of Luni River in
Rajasthan, a flashy flood in 1944 resulted in rapid rise of
water level of 24m in 2 days. In case of the Bandy, its
tributary, flood level rose even faster by 48 m in the same
period. The resulting afflux upstream of the bridge was about
1.83 m, which caused breaches in left approach and damage
to three end piers on the left flank.

Land slides some times cause natural damming.

When such dams burst, a devastating flood wave can be
generated as in case of Alaknanda River in July 1979. The
example of Sutlej River in August 2000 in Himachal Pradesh
has already been cited as recent example of such floods.
Photo 1.2 shows a damaged suspended bridge on river Ravi
located about 25 km downstream of Rampur town, during the
same flood. The suspension wires and abutments only are
left out after the floods. The normal procedure of arriving at
the design flood for a rivarian structure cannot cater for such
contingency. Such a possibility cannot, however, be
completely ignored. Suitable increase in design discharge
may, therefore, be made based on past experience of natural
dam bursts caused by landslides in the catchment under
consideration or adjoining catchments.

1.5.2 Virgin Rivers

Some of the rivers flowing in arid regions of

Rajasthan and Kutchh have no outfall in the sea nor do they
join any other stream. Such rivers, after traversing some
distance, lose all their waters by percolation and evaporation.
Rivers of this type are called Virgin Rivers. The example of
three main rivers of Kutchh can be cited for this purpose.

Rivers Banas, Khari-II and Saraswati originate in

Rajasthan, and travel through Kutchh. Normally, these rivers
experience very little or no flows even in monsoon season.
However, occasionally the rivers experience very flashy
floods. Though, Narmada Main Canal crossings for these
rivers have high design discharge (28615 cum/s, 9101 cum/s,

8
and 11862 cum/s respectively), most of the discharge is
known to vanish by the time they reach to the Rann of
Kutchh. In fact, the rivers are nearly non-existent within a
short distance of 20-25 km from the Canal Crossings.

1.5.3 Himalayan Rivers

Himalayas are geologically young and the nature of

rock formation is of sedimentary origin. Intensity of rainfall is
heavy. The region is susceptible to severe earthquake
shocks in addition. All these factors are responsible for big
and numerous landslides and enormous yield of sediment
from the catchments of Himalayan rivers. Thus, torrential
floods and heavy sediment load are the characteristics of
Himalayan Rivers. In addition to monsoon floods considerable
snow melt discharge makes these rivers carry perennial flow.

After debouching into the plains, all Himalayan Rivers

join the Indus, the Ganga or the Brahmaputra river systems.
These river basins are formed of deep alluvial deposits and
hence all watercourses in these basins belong to the
category of alluvial streams.

1.5.4 South Indian Rivers

In contrast to Himalayan geology, the South Indian

peninsula is geologically much older and formed of igneous
rock formations except at the outfalls of the rivers into the
sea where Mahanadi, Godavari, Krishna and Cavery deltas
are formed. Sediment yield and concentration are accordingly
smaller. These rivers belong to the category of incised rivers.
The term ‘incised’ characterises channels generally formed in
the process of degradation. The bed and bank material is,
therefore, different from the sediment transported in such
rivers most of which comes from the catchment due to
denudation and soil erosion. The bed and banks of the river
are usually highly resistant to erosion. The river systems in
South India are older and more stable. Hence, tendency for
shifting of river courses and for aggradation or degradation is
insignificant.

9
 Chapter 2

BEHAVIOUR OF RIVERS
A bridge engineer needs to be well conversant with the
river behaviour, which is largely influenced by the form or shape
of the river alignment in plan. Plan form of a river is designated by
the term “Channel pattern” which can be either meandering,
braided or straight.

Photo 2.1 (A) : (Top) Showing Brahmaputra River at

Guwahati. Straight course of the river is due to hillocks
located on both banks. (Bottom) Showing straight chhanel
of Kalang River, Assam

10
 Straight courses of rivers are not met with very frequently.
Leopold and Walman (2.1) have observed that they are more of an
exception, the straight lengths not obtaining for more than ten
times their widths.

Photo 2.1(A) shows more or less straight river courses.

In one case, it is due to the movement of channel restricted by
the hills. In the second case, the straight length is natural.
However, it can be seen that it might be a temporary phase,
supporting the above statement made by Leopold and Walman.

2.1 MEANDERING PATTERN

River courses normally follow tortuous alignment. Such

curvilinear courses are termed meanders. River meanders are
rarely regular and uniform in shape. The different shapes frequently
found in nature are — (i) Normal (Sinusoidal) (ii) flat (iii) acute or
intense or sharp, etc. Some of these are shown in Photo 2.1 (B).

Regular meanders-Narmada Regular flat meanders-Dangori

River at Hoshingabad, M.P. River,Assam

Regular sharp meanders- Regular and intense

Narmada River at meanders-Dangori
Hoshingabad, M.P. River,Assam

Photo 2.1 (B) : Examples of river meanders

11
Photo 2.1 (C) : Examples of complex type of river channel

Some complex type of river meanders are shown in Photo 2.1(C)

Some times the meanders change their shape or move

downstream. Such meander characteristics have important
bearing on the performance and safety of bridges and hence
warrant study.

The topic of remote sensing has been discussed in the

next chapter, the same may be referred to for further
understanding. However, the reader might not feel any problem
in understanding the different shapes of the river channels as
seen from the imageries.

2.1.1 Meander characteristics and causative factors

Meanders have been attributed to different factors by

different investigators. The more recent and widely accepted theory
is that formation of a meandering course is nature’s way of
minimizing variance of river parameters and the rate of expenditure
of energy. (2.2) (2.3)

12
 Some meanders are very flat while others very sharp or
acute. These and all possible shapes in between can be depicted
schematically using arc of a circle as shown in Fig. 2.2. River
meanders are rarely so regular and uniform in shape. The
conceptual model using circular arc, however help in quantitative
assessment of the behaviour of meanders. Definition sketch giving
symbols is presented in Fig. 2.3.

Shape of the meanders can be defined by the ratio of

length along the river to straight length along the valley, LR/LV.
For understanding how shapes of meanders with different values
of LR/LV can be developed. It is helpful to study the process of
bend erosion. In bend flow super elevation is obtained along the
concave bank which results in formation of secondary transverse
flow. Effect of this transverse flow is to shift the zone of high
velocity and maximum depth to the concave side and render the
side slope of the bank stipper. The river bank then becomes
unstable and fails and progressively recedes.

When the river section is narrow and deep, shifting of

the zone of higher velocity and bigger depth towards the concave
bank occurs earlier within a small bend angle and near the apex
of the bend. In the process, the meander curvature accentuates.
On the other hand, in a river with wide and shallow section, the
shift of high velocity and big depth of flow to the concave bank
requires longer travel distance and hence the location of maximum
depth and steepest side slope in a bend occurs downstream of
the apex point. Bank erosion at this location makes the curvature
flatter. Thus wider and shallower the cross section of the river,
the meander curvature becomes flatter. Meanders of different
tortuosities thus appear to be generated on account of differing
cross sectional shapes of the rivers.

2.1.2 Cut offs and movement of meanders

Because of progressive bend erosion in narrow and deep

rivers, the meander shape changes but not the location. With
continued erosion, a cut off occurs across the neck of a hairpin
bend as shown in Photo 2.4. The original course of the river then
turns in to a lake. Due to its typical shape, these lakes are
called Oxbow lakes. An example of such lakes is also shown in
Photo 2.4.

13
14
 Photo 2.4 : Showing development of cut-off observed in
Gango River, Assam and formation of Oxbow lakes

Relations have been evolved to arrive at the likely ratio of

(LR / LV) beyond which a cutoff is likely to develop. However, it
may be added here that such relations should be taken as only
tentative. There can be many other factors, which could act against
the development of cutoff. Photo 2.1(C) is one of such examples
where the cutoff has not developed.

Natural cut off increases the river slope and accelerates

the velocities in the upstream reach and thus brings about

15
formation of more cutoffs in its wake. On the other hand, in the
case of wide and shallow rivers, the meander shape does not
change but the whole meander train moves downstream. This
downstream travel of meanders results in cyclic changes of the
meander position within Khadir at any fixed location. In the case
of Ganga river at Mokamah and Mansi the meander was known
to occupy the same position either at the south or at north edge
of Khadir after a period of about 70 years as shown in Fig. 2.5.

Behaviour of meanders with flat and acute bends being

different, it is important to know the sinuocity 'LR/LV' associated
with these two shapes. For semi circular shapes, value of LR/LV
is generally less than 1.57, while in case of acute bends, LR/LV
is normally more than 1.57.

2.1.3 Meander relationships

Bigger the discharge, the meander length and belt may

be expected to be also bigger. From Lacey equation,
P = 4.836 Q½ , it is logical to presume that the parameter discharge
‘Q’ in cumecs can be expressed in terms of the parameter width
‘W’ in m which nearly equals the wetted perimeter ‘P’ in m in
case of rivers. Further more, when ML and MB are individually
related to Q, the ratio ML/MB should also permit quantitative
relationship. Working on such reasoning, various investigators
have attempted and evolved meander relationships. In India the
Inglis relationships are widely known. Schumm, Carlston, Leopold
and others in the USA, and by Ackers and Charlton in U.K have
also derived meander relationships. It was seen before that in
case of rivers with wide and shallow cross sections, the meanders
achieve flatter curvature and vice versa. It was also observed that
narrow and deep rivers have generally associated characteristics
of steeper slope and coarser bed material. A relationship for
tortuosities can, therefore, be developed between these relevant
parameters(2.4). All relationships mentioned above are presented
in Table 2.1(2.5).

16
 A
RS
SA F

HA
M RO
AS M

SA
TI
PU

TO
R
FR
OM A
RI SI
SA A AN
MA AG M
ST
IP KH
UR

NI
BARAU

90
19

19
TO
00

1969 KA
TI
HA
1969 R

17
1936

E
19

IDG RA
J
19
36

BR END
RA
19 9
69

REFERENCES :
1. COURSE OF R GANGA IN 1969 SHOWN THUS
2. COURSE OF R GANGA IN 1936 SHOWN THUS
3. COURSE OF R GANGA IN 1990 SHOWN THUS
Fig.2.5 :Cycle changes in Ganga river at Mansi
 Table 2.1

Sr.No. Investigator Year Relationship Data Remarks

1. Ferguson 1863 ML = 6W Ganga

2. Jefferson 1902 MB = 17.6 W American and European Rivers

3. Inglis 1939 ML = 49.63Q½ Shaw’s data of Orissa rivers in For 16 rivers in Flood Plain
India
MB = 14.0 W Bates data of American rivers For rivers in Flood plains

MB = 30.8 W Bates data of American rivers For Incised Rivers

ML/MB = 0.35 Jefferson data For Incised Rivers

18
R = 37.385 Q½ Jefferson data For Incised Rivers

ML/MB = 0.42 Jefferson data For rivers in Flood Plains

R = 25.358 Q½ Jefferson data for rivers in Flood Plains

4. Leopold et al 1964 ML = 11.03 W1.01 50 streams ranging from

models to large rivers
MB = 3.04 W1.1

ML = 4.59 R0.98

5. Prus Chacinski ML = 15.0 W — European Rule of Thumb

 Sr.No. Investigator Year Relationship Data Remarks

6. Ackers & 1970 ML = 61.19 Q0.467 Model data

Charlton
ML = 34.11 Q0.505

ML/MB=1.80 Q-0.038

7. Schumm 1963 LR/LV=3.5(W/D)0.27 M = % of silt and clay in the

-1.08
perimeter
W/D = 225 M
1967 Data on 47 Qm = mean annual
ML = 193.545 Qm0.34/M0.74 channels discharge

19
8. Chitale 1970 LR/LV=1.429(m/D)-0.077
S*-0.052 (W/D)-0.065 S* = S x 104
MB/W = 48.299(m/D)-0.050 Data of 42 rivers
S*-0.453 (W/D)-0.471 m = size of bed material
LR/LV=1.145(MB/W)-0.134

ML = meander length in m, W = channel width in m, MB = meander belt in m,

Q = Discharge in m3/sec, R = Radius of curvature in m, D = flow depth in m.

2.1.4 Engineering Implications of Meanders

In river engineering, meandering processes described

above have vital implications. The phenomenon of progressive
movement characteristic of relatively flat meanders can affect
working of engineering structures like bridges, water intakes,
diversion works etc. Movement of meanders results in cyclic
changes in the position of main channel within the Khadir as
shown in A of Fig. 2.6.

20
 These changes are not local but are governed by
changes occurring in upstream meanders. Nodal points occur in
rivers where natural constraints in the form of non-erosive strata
restrict downstream movement of meanders. In Ganga River, the
Kasmar bluff on the north bank has thus developed a nodal point
at Patna as seen in Fig. 2.7.

FIG. 2.7 : Superimposed courses of

RiverGanga at Patna, Bihar
showing nodal points

Nodal points provide good locations for constructing

engineering structures since the river width here is narrower and
the channel position more stable. Bridges and barrages provide
artificial constraint to free movement of meanders in the
downstream direction. This results in the meander loop getting
squeezed and distorted immediately upstream of the structure
posing threat to the safety of approach banks as indicated in B
of Fig 2.6. Progressive erosion in acute bends results in
development of cut offs across the neck, schematically shown
in C of Fig. 2.6.

If structures are either proposed or already constructed

in such rivers, it is likely, that occurrence of cut offs will affect
them. In extreme cases, a cut off can render an existing structure
obsolete and redundant by short-circuiting the river reach wherein
the structure is sited.

21
2.2 BRAIDING PATTERN

Formation of braiding pattern is popularly attributed to

heavy sediment load in a river having a wide and shallow cross
section leading to sediment deposition during falling flood when
transporting capacity is reduced quickly and appreciably.
Implication of the channel section being wide and shallow is that
non-uniformity in flow distribution; sediment size and sediment
transport across the section is then obtained more easily. These
conditions favour formation of islands and braiding pattern
emerges. Brahmaputra River is a notable example of braided
river as shown in Photo 2.8. Photo 2.8 shows two types of braided
planforms, namely, (i) interlaced type - where number of channels
intersect each other and (ii) island type - where the channels
rarely cross each other. Thus, long islands are formed.

The islands in braided rivers travel downstream. Discharge

carrying capacity also changes over the period causing changes
in the channel alignment. Such changes cause bank erosion.
Such activity is shown in Photo 2.9.

The braiding pattern when accompanied by aggradation

results in lateral shifting of the river course rendering the river
unstable both in plan and elevation. The Kosi in north Bihar was
a well-known example of this type of river prior to construction of
flood embankments. In a period of 200 years, the river shifted
over 122 km from east to west as shown in Fig. 2.10.

Such conditions are very unfavourable for engineering

structures. The aspect of stability is, therefore, required to be
examined when a project is conceived in a braided river.

Many braided streams have been trained to provide a

single deep meandering course by means of lateral constriction
works. Portion of the Danube River in Hungary trained in this
fashion is shown in Fig. 2.11.

Such a single course has more capacity for conveyance

of water and sediment than a braided river. This kind of river training
is, however, possible when tendency for significant aggradation
is absent.

22
Interlaced type braided channel – Brahmaputra River, Assam

Island type braided channel – Luhit River, Assam

Combined type braided channel – Brahmputra River, Assam

Photo 2.8 : Examples of braided river channels

23
 Photo 2.9 : Progressive development of channels,
shifting of islands and consequent bank erosion in river
Brahmaputra, Assam

2.3 BED FORMS

These are deformations of river bed caused by the flow.

In a laboratory flume such deformations can be generated by
progressively increasing the velocity. The sandy bed then develops
bed forms in the sequence of no movement, threshold movement,
ripples, dunes, transition and anti dunes schematically indicated
in Fig.2.12.

Threshold movement occurs when the bed shear or

bottom velocity achieves the critical value required for movement
which is roughly given by the equation Tc = 0.06 (rs - r) d50 wherein
Tc is the critical shear stress in kg/cm2, rs is the specific weight
of sediment in kg/cm3, r is the specific weight of water in kg/cm3
and d50 is the size of sediment grains in them such that 50 percent
of the particles are finer than this size.

Ripples are small bed forms obtaining after threshold

movement stage and have a characteristic of downstream
movement. Wavelength is normally less than 0.3 m and wave

24
25
26
Fig.2.5 : Training of Danube river in Hungary
 (a) TYPICAL RIPPLE (e) PLANE BED
PATTERN, F << 1. & d < 0.4 mm.

DECELERATION ACCELERATION
ACCELERATION WEAK BOIL

(b) DUNES WITH RIPPLES (f) STANDING WAVES,

SUPERPOSED, F<<1. F > 1.

BOIL BOIL
F = FROUDE NUMBER

(c) DUNES

(d) WASHED OUT DUNES

OR TRANSITION.

Fig.2.12 : Schematic depiction of some of the

types of bed forms

27
height less than 3 cm. For ripple formation the size of bed material
needs to be less than 0.4 to 0.6 mm. Dunes have wave length
and wave height greater than ripples and these are dependent on
flow characteristics. They also move downstream and are
associated with sub critical flow regime (with Froude no. less
than 1). As the flow approaches critical condition, when Froude
Number (V/ gD ) approaches 1 (where V is velocity in m/sec, D
depth in m), dunes start getting washed out due to high velocity.
This is the transition bed form, which is some times called flat
bed. With Froude Number equal to or greater than I, antidunes
are formed whose wavelength is roughly equal to 2p V2/g where
V is the velocity in m/s and g is the gravitational acceleration.
Their height depends on flow characteristics. Antidunes are sand
waves of sinuous shape in phase with gravity water surface waves
and may move upstream, downstream or remain stationary
depending on flow characteristics.

Against various bed forms described above observed in

the laboratory flume, the bed forms normally met with in a river
are ripples, dunes and washed out dunes. Important
characteristics of bed forms having wide implications in river
engineering are their magnitude and mobility. Table 2.2 gives
characteristics of bed forms in Brahmaputra river in its lower
reach in Bangladesh(2.6).

Sand waves of 7.55 to 15.25 m height and 183 to 915 m

length were observed moving about 204 m in a day down the
river. The depth of flow at this stage varied from 6.08 to 9.15 m.
Similarly in the case of Mississippi river between New Orleans
and Old river, a distance of 320 km, it was observed that wave
heights of as much as 9.15 m were obtained in 24.4 m depth of
water(2.7). These sand waves were found to change systematically.
They became larger with increase in discharge and smaller with
decrease in discharge.

Implications of such bed forms are varied. Movement of

big sand waves across bridge piers or intake wells may at one
stage bury them in sand while at the other stage expose them
excessively rendering them unsafe. Secondly, measurement of
river discharge by velocity-area method often consumes a day.

28
 Table 2.2 : Characteristics of Bed Forms in the lower
Reach of Brahmaputra River

Characteristics Small Size Big Size Dunes Sand

Ripples Ripples Waves

Range of Wave Upto 0.3 m 0.3 to 1.5 m 1.5 to 7.6 to

Height (WH) 7.6 m 15.2 m

Range of Upto 1.5 m 3.0 to 152 m 42.7 to 183 to

Wavelength 488 m 914 m
(WL)

Range of bed 1:5 to 1:20 1:6 to 1:100 1:30 to 1:25 to

form index 1:60 1:100
(WH/WL)

Maximum
amount of
movement in 24 — 246 m 159 m 640 m
hours period

Average amount
of movement in 3m 122 m 67 m 204 m
24 hours period

Big sand waves moving across the measuring section can,

therefore, affect accuracy of discharge measurement and cause
scatter in the plotting of observed data in preparing stage-discharge
curves, which are also termed rating curves. One more aspect is
about Rugosity Coefficient in a flow formula, which depends largely
on bed roughness governed by bed form. Estimation of velocity
using a Manning type formula, therefore, requires proper
assessment of bed form and associated value of Rugosity
Coefficient.

Relationship defining type of bed form as a function of

hydraulic radius R in m, slope s, mean velocity V in m/sec and
grain size in mm was evolved by Simons and Richardson as
shown in Fig. 2.13.

29
 40
30

20
UPPER REGIME
F>1
10.0
TRANSITION
8.0
6.0
F = FROUDE NUMBER
4.0
3.0

2.0
DUNES
F>1
1.0
0 0.2 0.4 0.6 0.8 1.0 1.2
MEDIAN GRAIN SIZE IN MILLIMETRES

Fig. 2.13 : Relation of stream power and median grain

size to form of bed roughness

Recent studies suggest that the lower regime of bed

forms will occur when the ratio
4
V
2 1/2 2/3
g D d 50

is less than 1 x 103, that the upper regime of bed forms will occur
when the ratio is greater than 4 x 103 and that the bed will be in
transition if ratio is in between these values. In the above ratio, V
is the mean velocity in m/sec, D is the mean depth in m and d50
is the median grain size in mm.

2.4 RIVER FORMULAE

Schumm(2.8) dealt with interdependence of various river

parameters in a qualitative way. In a river system, the parameters
water discharge Q, bed material load Qs, bankful stage width W,
corresponding average depth D, slope S, meander wave length

30
ML, ratio of channel length to valley length P etc. all interact and
change, results in changes in the remaining parameters. Using
a plus exponent for increase and minus for decrease, Schumm
indicated the nature of change brought about in channel
morphology by change in water discharge and sediment load.
The associated changes were qualitatively expressed by him in
the following form.
+ + + - - -
+ W D ML _ W D ML
Q ≅ _ Q ≅
S S+
+ + + _ - - -
+ W ML S W ML S
Q ≅ _ _ Qs ≅
s D P D+ P+

Lane(2.10) presented a qualitative relation

Qs . d50 ≅ Q.S

wherein d50 is the size of the sediment as defined earlier.

Quantitative proportionalities were developed by

Langbein (2.11). He advanced the hypothesis of uniform distribution
of changes among the dependent variables and evolved structure
of an exponential relationship between dependent and
independent variables. By considering discharge to be an
independent parameter he determined the exponents in the
formulae of the type.

Width = coefficient x discharge

The proportionalities so obtained by him were

W α Q0.5
D α Q0.33
S α Q-0.17
V α Q0.17
Empirical formulae for primary variables such as width,
depth and slope were evolved by Lacey, Blench and several other
authors. More important of the formulae are listed in Table 2.3(2.9).

31
 Table 2.3 : River Formulae

S.No. Author Formulae

Width Formulae
1. Lacey W = 4.836 Q0.5
2. Blench W = Fb0.5(Fs-0.5)Q0.5
3. Nixon W = 2.988Q0.5
4. Pettis W = 4.438 Q0.5
5. Statistical W = 1.434 Q0.949 D-1.237
6. Statistical simplified W = 1.6Q D-1.5
Depth Formulae
7. Lacey R = D = 0.473Q0.33f-0.33
8. Lacey R = D = 1.34q0.67 f-0.33
9. Blench D = Fb-0.67Fs0.33Q0.33
10. Nixon D = 0.539 Q0.33
11. Pettis D = 0.635 Q0.30
12. Statistical D = 1.339 Q0.767 W-0.808
13. Statistical Simplified D = 3.6 Q0.8 W-1.0
Slope Formulae
14. Lacey S = 0.000309 f-1.667 – Q-0.167
15. Blench S = 0.00684 Fb0.833 Fs0.833 Q-0.167
16. Lane S = 0.0042 Q0.25
17. Statistical S x 104 = 232 Q7.200 W-7.767 D-9.600
18. Statistical Simplified S x 104 = 7.5 M0.5
Velocity Formulae
19. Lacey V = 4500 RS for bed material
size <0.2mm
20. Lacey V = 44.59 R0.75 S0.50 for bed material
size between 0.2 and 0.6 mm.
21. Lacey V = 10.81 R0.67 S0.33 for bed material
size between 0.6 and 2.0 mm
22. Lacey V = 6.084 R 0.6 25 S0.25 for bed material
> 2 mm.
23. Nixon V = 0.6213 Q0.17
24. Pettis V = 0.4974 Q0.2
Formulae for Bed Material size
25. Statistical M x 104 = 571 Q14.31 W-15.103 D-18.379

32
Notes: (i) All formulae are in M. K. S. units except that
of Blench
(ii) In Blench formulae Fs was assumed as 0.1
and Fb = 1.9 mO.5 for size of bed material in
between 0.2 and 2.00 mm.
(iii) Blench formulae were evolved on basis of canal
and river data vide Reference 2.12. Lacey
formulae were evolved on basis of canal data
vide References 2.13 and 2.14. Nixon formulae
were based on data of U.K. rivers vide Reference
2.15. Pettis formulae were obtained using data
of U.S.A. rivers vide Reference 2.16. Statistical
relations were worked out using data of Indian
rivers vide Reference 2.17.

These empirical relationships are applicable to rivers

having same characteristics as of the streams, data of which
were considered in their derivation.

River formulae at Sr. Nos. 5, 12, 17 and 25 in Table 2.3

can be used to gain qualitative as well as quantitative design
data in varied types of river regime problems(2.17). In such
application, the physical constraints in any given problem are
important. The parameters involved are generally discharge Q,
width W, depth D, slope S, bed material size M and sediment
load Qs. Not all of them may be free to change. Frequently some
are changed purposely to effect the desired remodeling of the
river. For instance, in order to improve navigation depths, river
training is practiced by contracting the width. The ratio of original
width of the river W1 to the constricted width W2 can take various
values depending on the extent of contraction. Formulae 5, Table
2.3 then takes the form.
0.95 −1.24
W1 ⎡ Q1 ⎤ ⎡D ⎤
= X ⎢ 1⎥
W2 ⎢⎣ Q 2 ⎥⎦ ⎣ D2 ⎦

Putting Wl/W2 = Wr, Q1/Q2 = Qr, D1/D2= Dr

S1 / S2 = Sr and M1 / M2 = Mr and substituting Qr = 1,
formulae 5, 12, 17,25 in table 2.3 give the following relationships

33
 Dr = Wr-0.8
Sr = Wr0.24
Mr = Wr-0.25
Vr = Wr-0.19

Constriction in river width thus results in increase in river

depth, flattening of slope, coarsening of riverbed material and
increase in mean velocity of flow.

It is general experience that combining several channels

of a stream into one increases its transporting capacity for
sediment load. It is possible to assess changes in width, depth
and velocity resulting from conversion of multi-channel river into
a single channel one, the operative constraints being slope and
bed material size remaining practically unaltered. The width and
depth relations obtained under these conditions were:

Wr = Qr1.66
Dr = Qr -0.57
Vr = Qr -0.03
Similarly effect of Constructing flood embankments on
depth and velocity can be evaluated. The most general case is
the one where all parameters are free to change without
constraint. In this case the relations obtained were the following.
Wr = Qr0.50 (Assumed)
Dr = Qr 0.36
Sr = Qr-0.27
Vr = Qr 0.14
Mr = Qr 0.10
Thus the changes in river parameters are interdependent
and when the ratio of variation in anyone of the parameters is
known, the corresponding ratios for the other parameters can be
derived.

34
REFERENCES

2.1 Leopold, L. B. and Walman, M.G., ‘River Channel

Patterns: braided, meandering and straight’, U.S.
Geological Survey, Professional Paper 262 B, 1957.
2.2 Langbein, W.B. and Leopold, L.B., ‘River meanders
theory of minimum variance’, U. S. Geological Survey
Professional papers, 422-H:H1-H15 1966.
2.3 Yang, C.T., ‘On river meanders’, Journal of Hydrology
13 (1973), North Holland Publishing Company,
Amsterdam.
2.4 Chitale, S.V. ‘River channel patterns’, journal of
Hydraulics Division, American Society of Civil Engineers,
January 1970.
2.5 Chitale, S. V., ‘Theories and relationships of river channel
patterns’, Journal of Hydrology, 19 (1973), North Holland
Publishing Company, Amsterdam, Netherlands.
2.6 Coleman, J.M., ‘Brahmaputra river, channel processes
and sedimentation, Sedimentary Geology’, Elsvier
Publishing Company, Amsterdam, Netherlands, Volume
3, August 1969.
2.7 Carry, W.C. and Keller, K.D., ‘Systematic changes in
the bed of alluvial channels’, Journal of Hydraulics
Division, American Society of Civil Engineers, Volume
87, No. Hy 3, Proceedings Paper 2816, May 1961.
2.8 Schunm, S. A. ‘River Metamorphosis’, Journal of
Hydraulics division, American Society of Civil Engineers,
January 1969, pp 255-273.
2.9 Chitale, S. V. ‘River Formulae’ Irrigation and Power,
Journal, Central Board of Irrigation and Power, New
Delhi, April 1980 pp 205-218.
2.10 Lane ,E. W., ‘The importance of fluvial morphology in
hydraulic engineering’, Proceedings American Society
of Civil Engineers, Vol. 81, No 746, 1955.
2.11 Langbein, W.B., ‘Geometry of river channels’ Journal of
Hydraulics Division, American Society of Civil Engineers,
March 1964, pp 301-312.

35
2.12 Blench, T., ‘Regime behaviour of Canals and Rivers’
Butterworths Scientific Publications, London U.K. 1957.
2.13 Lacey, G., ‘Stable channels in alluvium’ Paper No. 4736,
Minutes of Proceedings of Institution of Civil Engineers
London U.K. Vol. 29, 1930, pp 259-384.
2.14 Lacey, Q and Pamberton, W, ‘A general formula for
uniform flow in self formed alluvial channels’ Proceedings
of the Institution of Civil Engineers, London, Part 2,
September 1972.
2.15 Nixon ,M., ‘A study of bankful discharges of rivers in
England and Wales’, the Institution of Civil Engineers,
Proceedings, February, 1959, Vol. 12, Session 1958-
59, Paper No. 6322.
2.16 Pettis, C. R., ‘Discussion published in American Society
of Civil Engineers Transactions 1937’, pp 149 to 152 on
Paper No. 1957, ‘Stable Channels in erodible material’
by E.W. Lane.
2.17 Chitale, S.V. ‘Sympathetic changes in river regime
Proceedings, Institution of Civil Engineers, London U.
K. Part 2, 1977’13, September, pp. 613-623

36
 Chapter 3

REMOTE SENSING AND ITS APPLICATION

FOR BRIDGE ENGINEERING

As a part of National Natural Resources Management

System, India launched Remote Sensing Satellite IRS-1A, way
back in the year 1984. Since then, there is long series of satellites
like IRS-1B, 1-C, P4, P6, etc. The satellites have provided handy
and effective tool for continuous and operational remote sensing
data for the management of natural resources. Aligning the new
routes, maintenance of the existing roads and railways,
construction and maintenance of bridges, providing protection to
the bridges, etc are integral parts of the communication system,
in which, remote sensing has a significant share.

Use of remote sensing for application in river engineering

is building up at a rapid pace. A bridge engineer is required to
understand the hydraulic characteristics of river and the river
behaviour in the vicinity of the bridge, etc before taking up the
construction of a bridge. In view of safety of an existing bridge,
he is required to design suitable river training measures. For
many such aspects, study of satellite data of the past and
present, collected through the satellites, has proved highly useful
and dependable tool.

Changes in the river channel alignment and size,

developments or decay of shoals, chars, islands, etc take place
during floods. After studying the past data, accurate and
dependable predictions of river behaviour can be made for long
reaches covering upstream and downstream of the bridge.

Other information like existence of palaeo channels, low

lying areas, reaches under active erosion, damages due to over
bank flows, breaches in the embankments, damages to the
existing hydraulic structures etc can be identified using satellite
imageries.

Before discussing different aspects of remote sensing

relevant to the bridge engineer, it is necessary to take an overview
of the remote sensing technique.

37
3.1 REMOTE SENSING TECHNIQUE

Remote sensing is a science or an art of obtaining the

information about an object, area or a phenomenon through
analysis of data acquired by a device that is not in contact with
the object, area or the phenomenon under investigation. For
example, reading is an art of remote sensing where the eyes act
as sensors, which respond to the light reflected from the page of
written material and collect the data, which is analysed in the
brain to obtain the information.

3.1.1 Satellites

Satellites used for remote sensing are of two types, viz,

geostationary and polar orbiting. Geostationary stationary
satellites are located at the equator at a very high altitude. Their
speed and direction are so adjusted that a viewer on the earth
sees them as stationary. Polar orbiting satellites rotate around
the earth following a path from North to South Pole. Fig 3.1 shows
the two types of satellite orbits with reference to the earth. The
path and speed of polar satellites are so adjusted that they travel
along the same path and at the same time at pre-determined
regular interval.

Fig 3.1 : Two types of orbits followed by Indian satellites

3.1.2 Scanning process

The difference between a photographs and imageries

38
needs to be understood first. A photograph covers a specific area
within its view and a picture is taken covering the full area at a
time. Everybody is familiar with this process. A satellite observes
and records the data line by line, which is made up of many
points. The point-by-point observations of the intensity of light
are taken in digital form. Fig 3.2 shows the movement of satellite
and the process of observation in schematic way. The process
can be compared to a scanner where the source of light travels
over the document, collects, and stores the data in digital form.
Therefore, a satellite takes certain time to capture the data
covering the specific area of interest.

Fig 3.2 : Method of observation by satellite

The collected digital data can be converted and printed

on photographic paper to form a "photograph" called in the
technical term as "imagery". This process can be compared to
a Xerox machine, where a document is "scanned" and the data
is printed on a paper.

Aircrafts and satellites are the other most commonly

used methods for data acquisition from the earth surface. Aerial
photography, i.e. remote sensing of data using aircrafts, took up
shape with the development of aircrafts and its suitable remote
sensing cameras. Aerial photography has the advantage of
flexibility of operations and higher resolution.

39
3.2 ADVANTAGES

Remote sensing has gained quick popularity due to the

following advantages:-

i. Synoptic Coverage : Depending upon the area of

interest, areas covering few hundred sq m to thousands of sq km
can be suitably collected and studied. Photo 3.3 shows an
example of the advantage of satellite data over index plans due
to synoptic coverage. Satellite imageries of river Brahmaputra
covering a length of about 150 km and about 10 km are shown
on the right. Corresponding index maps prepared using other
available maps are shown on the left side.

Photo 3.3 : The synoptic coverage. The right side

photographs are the satellite imageries covering about 150
km (upper) and about 10 km (lower) of river length. Index
maps are on the left.

Bridge engineer may consider this aspect equivalent to

the plane table survey plan, drawn at a suitable scale. In this
view, satellite imageries serve the same purpose, but with more
accuracy and dependability.

ii. Repetitive coverage : Due to repeated passes of the

satellite over the same reach, data is repeatedly acquired.

40
Therefore, comparisons of a dynamic phenomenon, like river
channel changes, can be studied easily. Photo 3.4 shows an
example of river channel alignment observed at an interval of four
years. Analysis of river channel changes during the period can
help to assess the changes and estimate the likely changes in
future.

Photo 3.4 :The repetitive coverage. Imageries taken at an

interval of four years show the river channel changes.

iii. Inaccessible area coverage : Marshy lands, hilly

areas, etc where ground surveys could be very difficult, remote
sensing technique can provide reliable and correct information.
Photo 3.5 shows a river channel observed in the imageries and
in the field. Difficulties in the field observation can be overcome

Photo 3.5 : Showing inaccessible river reach and

corresponding satelite image.

41
by the use of satellite data.

iv. Amenability to computer processing : Due to the

digital data collected and stored in the form of a matrix, the same
becomes amenable to processing on computers. By following
certain procedures, reliable identification of features and
superimposition of the same to assess the changes, etc can be
easily and accurately done by computer processing. Handling of
large volume of data, comparison of data having different
resolutions, etc becomes easier with the aid of computers.

3.3 CHARACTERISTICS OF DATA

Sensors are devices to make observations, which have

sophisticated mechanism for taking observations. The detectors
are designed to have specific electromagnetic properties. The
detectors help to observe the desired and meaningful information
from the earth's surface. Important parameters of sensors are
the spatial, spectral and radiometric resolutions.

i. Spatial Resolution : Spatial Resolution is the length

and width of the smallest object that can be discriminated by
sensors. This is commonly known as "pixel". Smaller the size of
pixel, greater is the volume of data covering the same area. Area
coverage and pixel determine the scale of imagery.

The Indian satellites acquire the data in three resolutions,

as below.

a. Wide Field (WiFF) sensor has a ground resolution

of about 188 m and width of observation (Swath) is
about 810 km.
b. Linear Imaging Self Scanning (LISS) sensor has
ground resolution of about 23.5 m and swath of about
141 km.
c. Panchromatic camera (PAN) has a high resolution
of about 5.8 m and swath of about 70 km.

Clarity in observation due to higher resolution, i.e. smaller

pixel size can be seen in Photo 3.6. The clarity of houses, roads,
trees, etc of the same reach covered by the data of different
resolution is evident.

42
Photo 3.6 : Showing imageries of low and higher resolution
covering the same area. Clarity of finer details is evident

ii. Spectral resolution : The sensitivity to different features

of earth is the most important aspects considered in the spectral
resolution.

The earth features can be divided broadly into three types,

viz, healthy green vegetation, dry bare soil and clear lake water.
It has been observed that each of these features has a clear and
distinct average reflectance curves in visible electromagnetic
range. However, in the nature, such features are rarely found in
their pure form. The vegetation, according to their environmental
conditions, have a range of reflectance depending upon the types,
their water contents, stresses, diseases, etc. Similarly, the re-
flectance from the soil changes according to its moisture con-
tents, soil textures, surface roughness, iron contents, organic
contents, etc. The reflectance of the clear water is highly influ-
enced by turbidity, algae contents, oil and other industrial / chemi-
cal wastes, etc.

Advantage of these characteristics is taken by taking

observations in different band frequencies of electromagnetic
waves within visible light range and infrared range. Different sen-
sors are assigned separate narrow bands. Different bands help
to give different identifications and classifications to the objects
viewed using the data. Photo 3.7 shows rivers and adjoining land
observed by different bands. Analysis of data showing the differ-
ences helps to understand the topography in a better way.

43
Photo 3.7 : Showing the river reach observed by different
bands

iii. Radiometric resolution : It is the capability of the sen-

sor to differentiate the reflectance from various objects on the
earth. For the darkest to brightest object on the observed, one
camera might show values from 0 to 32, whereas, the other from
0 to 256. This sensitivity depends up on the number of bits used
to express a number in digital data.

3.4 ACQUISITION AND PROCESSING OF DATA

The steps involved before supplying the remote sensing

data to the users are as follows :

a. Observations are taken using sensors like cameras,

which are mounted on suitable platforms like aircrafts, satel-
lites, etc. In case of satellites, the cameras, mounted in the
satellite looking down to the earth, make continuous observa-
tions. The observations made, in digital form, are transmitted to
the receiving stations on the earth. The data covering a specific
reach is called as imagery of that reach.

b. The receiving stations record the observations on a

suitable medium like photo papers, magnetic tapes, CDs, etc.

44
These also monitor the exact path, height, location, etc of the
satellite when the data is received.

c. The data received from the satellite is called raw data.

Distortions are possible in the observed data due to motion of
the satellite relevant to earth, changes in the altitude,
misalignments, position errors, curvature of earth, non-uniformity
of illuminations, variations in the sensor characteristics, etc. The
data, being digital, is amicable to corrections. The desired
corrections are made using computers.

d. The corrected data is supplied to the user in its digital

form for further analysis and interpretation using computers. The
digital data is supplied on magnetic media like floppies, data
cassettes, CDs, etc. The digital data requires computers having
suitable softwares and some experience and expertise for further
analysis and interpretation.

The supply can also be in the form of photo prints taken

on photo films using computer based photo write systems. The
photo print is normally black & white (B&W) when data of only
one band is required. When multiple bands are to be printed, the
colour of the bands is used for printing. The output prints do not
show “natural” colours. For example, the green colour of the
vegetation looks red in the print. Such multi-band printing is called
as False Colour Composite or FCC print.

3.5 STUDY OF RIVER MORPHOLOGY

Study of river morphology gives an insight in to the

changes in the rivers in planform caused due to inherent and
man-made instabilities. The changes in the river planforms can
be systematic / regular which can be predictable, as in case of
single channel meandering flow, or complex where effect of many
interactive forces result in irregular / braided plan forms. In case
of braided rivers, the assessment of cause / effect relationship is
slightly difficult.

3.5.1 Interpretation and analysis using remote sensing

Various characteristics of a river can be understood by

45
studying its planforms, their changes over periods, etc. Many of
these aspects are important for a bridge engineer. River layout
and planforms, like, meandering, braided, straight, etc can be
seen in a synoptic view of the imagery. Further sub-division of
the rivers like (a) sharp and flat, irregular and irregular, intense,
type of meandering rivers (b) island type and interlaced type braid-
ing rivers (c) straight but stable / unstable, etc can be easily
identified by the use of synoptic view of the imageries.

Detailed observations and analysis of the data can help

to identify many features on the ground which are relevant to the
study of rivers, like hills, rock outcrops, etc; hydraulic features
like old, dead and palaeo channels, oxbow lakes, dry channels,
flood channels, offtake channels, large and small tributaries etc.
Man-made structures like roads, railway lines, villages and towns,
etc; hydraulic structures like embankments, long spurs, barrages,
weirs, bridges with their appurtenant works like guide bunds,
approach embankments, etc can also be identified using B&W
and FCC prints in combination or separately.

Temporal data covering the desired period can help to

study the configuration of river channels and other hydraulic struc-
tures in the past. Analysis of the data helps to study the shifting
of river channels, development of shoals and islands, changes in
size of dry weather channels, bank line changes, nodal points in
meandering rivers, stable and unstable reaches, Khadir limits,
etc. The study in turn helps to decide the location for abutments,
waterway required from stability point of view, vulnerable reaches
which may require river training / anti-erosion measures for safety
of the bridges, etc.

When FCC imageries are used, a bridge engineer can

visually see the changes in the colours, variation in the colour
patterns in specific reaches, etc. Inspection of site helps to verify
the features which can help to assess the flood prone areas,
areas where shallow and stagnant water can be met with, den-
sity of forest in the bridge approaches, etc.

Once the features as observed on the ground and corre-

sponding features observed on the photo prints of satellite imag-
eries is understood, a bridge engineer can undertake the prelimi-
nary analysis and decide many aspects of the bridge param-

46
eters qualitatively. However, for reliable quantification is required,
computer analysis is recommended, wherein suitable softwares
and expertise in the digital analysis is necessary.

3.6 PROCUREMENT OF DATA

National Remote Sensing Agency (NRSA), of the De-

partment of Space (DOS) is located at Balanagar, Hyderabad -
500 037. NRSA is nodal agency for the distribution of the data
acquired by the Indian remote sensing satellites. NRSA has a
large data bank collected by Indian satellites since 1984 till date.
Data older than that is also available from the American satellite
Landsat. NRSA also has a large collection of data acquired by
other foreign satellites like NOAA, SPOT, etc. These data are
stored and supplied to the users according to the availability.
Except few areas critical from security point of view, all the data
available with NRSA is fully open for any user.

NRSA requires a requisition indicating the exact require-

ment of the imageries. A bridge engineer newly intending to use
the data may face difficulties in the procurement. Following steps
are suggested which can be followed for smooth procurement of
the correct data.

Based on the response of the users, NRSA modifies the

methods and adds new facilities in their procedures. The most
common procedure, which can be easily adopted by the new
user, has been discussed below.

3.6.1 Steps for the procurement

a. Initially, an introductory letter may be sent addressed to

The Head, NRSA Data Centre, National Remote Sensing Agency,
Balanagar, Hyderabad – 500 037. The letter should clearly indi-
cate purpose of studies, type of studies involved, area of inter-
est; expected resolution; period of data to be covered for each
imagery, scale of the photo print required, etc. Requisition form
may also be requested in the introductory letter. Details of these
items are discussed separately.

b. Based on the information, NRSA will browse the avail-

able data and issue “Proforma Invoice“ indicating suitable satel-

47
lites, dates of pass showing cloud free data, and cost of each
photo print. More than one date of pass might be suggested
within each of the period interest. In such case, the bridge engi-
neer should make selection of suitable date of pass.

c. Bridge engineer should duly fill the requisition form and

a Demand Draft covering 100% cost of the imageries should be
drawn and enclosed with the requisition form and sent to NRSA.
If more than one date is available in the desired period, alternate
dates of pass, in order of priority, may also be indicated in the
covering letter.

d. The desired data is normally made available within a

period of 3 to 4 weeks. Some times, NRSA faces problems in
retrieving old data, percentage of cloud cover for the desired area,
missing lines of data, etc. Under such circumstances, NRSA
will make use of the alternate dates specified in the covering
letter. If such alternate dates are not available, NRSA will come
back to the user for suggesting new period for the date of pass.
Fresh look at the suitable date of pass may have to be taken
under such conditions.

3.6.2 Details for the requisition

Explanation regarding the details to be filled in the req-

uisition form is given below for ready reference.

i. Period of Data

Normally, for the morphological studies of a river, dry

weather data is required. Morphological changes in a river take
place due to floods during monsoon season. Changes in the
river planforms, shifting of channels hardly take place during lean
period. Therefore, suitable data of lean period, i.e. from October
to December and January to May is always acceptable. The
exact period should be specified for clarity, viz, October 1986 to
May 1987; October 1991 to May 1992; October 1996 to May
1997, etc.

For full clarity, only cloud free data should be requisi-

tioned. Normally, sufficient choices for the date of pass are avail-
able in a specified year.

48
ii. Area of interest

The user ( i.e. the bridge engineer) is required to decide

and specify the area of interest. The area is indicated as a rect-
angle of sides parallel to north-south and east-west direction.
This can be specified as follows

a. Longitude and latitude of the lower left corner and upper

right corner of the rectangle

b. Survey of India map sheet Number can be indicated.

This is normally readily available with the bridge engineer. There-
fore, this can be the most common method to indicate the de-
sired area.

c. Alternatively, the longitude and latitude of the bridge can

be indicated. NRSA can treat this location as centre point of the
map sheet to be supplied. This facility is applicable for the high
resolution data if IRS 1C and 1D, P4, etc. These data are avail-
able from the year 1996.

d. Path and row number of the desired satellite can be

indicated. NRSA has prepared and published maps showing path
and row numbers of different satellites. For a new user, this may
not be easily available. Experienced user and the users who
process the data on computers are normally familiar with this
alternative.

iii. Reach to be covered

River reaches to be studied with the help of satellite

imageries upstream and downstream of the bridge depends up
on the likely effect extending upstream and downstream. There-
fore, the lengths can be different depending up on the type of
rivers, its horizontal and vertical stability, etc. In case of mean-
dering and stable rivers, the length can be three meanders up-
stream and two meanders downstream of the bridge. For braided
rivers, the length can be four to five times its maximum width in
the upstream direction and three times the maximum width down-
stream of the bridge.

From the point of view of safety of the bridge, if the analy-

49
sis indicate necessity of protection / training measures in the
upstream, then the length of study reach can be accordingly
extended in the upstream direction.

iv. Bands and resolution

Normally, the LISS data is found sufficient for the stud-

ies. However, for important structures at critical locations and
structures located in the urban reach, additional data with higher
resolution i.e. say PAN data can be procured covering the de-
sired river reach.

v. Number of photo prints

Bridge engineers would normally require data on photo

prints of bands 2, 3 and 4 for FCC prints, or of only band 4 for
B&W prints. For conducting preliminary studies, FCC data of
the latest available period is required, whereas, for conducting
studies in detail, data at an interval of 4-5 years till the latest
data is normally sufficient. Procuring the FCC data for the oldest
and latest imageries and B&W of band 4 for the other years is
normally sufficient for analysis of river behaviour. In addition to
the above, as discussed above, PAN data covering a specific
reach can be procured.

3.7 CONCLUDING REMARKS

The above discussions regarding remote sensing give

only introductory information. An attempt to procure the data
and / or a visit to NRSA, Hyderabad will definitely help to under-
stand the procurement and analysis of the information in a better
way.

The different parameters for the bridges, which can be

studied with the help of satellite data, are discussed at their
proper location in the subsequent chapters.

50
REFERENCES

3.1 “Remote Sensing and image interpretation”, T M

lillesand, R W Kiefer, Published by John Wiley & Sons Inc, New
York, Singapore, 1994.

3.2 Lecture notes on “Application of Remote sensing and

GIS techniques in water resources development and manage-
ment “ Edited by National Water Academy Khadakwasla, Pune
– 411 024.

3.3 “Satellite data products from NRSA Data Centre”, infor-

mation brochures and booklets published by NRSA, Hyderabad
from time to time.

51
 Chapter 4

PRINCIPLES OF RIVER TRAINING

AND PROTECTION WORKS
Determination of design discharge for a bridge is some
times required to be based on flood marks left during
unprecedented abnormal floods. A flow formula is then used to
estimate discharge corresponding to flood marks. The river section
is often composite comprising flow section with one or more
deep channels and an overbank area carrying spill flow. Allowance
for such variation is required to be made in application of the flow
formula, suitable procedure for which needs to be known.

Excessive constriction of waterway by approach banks

results in rise of water level which is termed afflux. Apart from
construction of waterway, afflux is also caused due to pier
obstruction. The extent of afflux is an important consideration in
fixing bridge waterway.

In the case of alluvial rivers, the normal Indian practice

is to provide constricted waterway for a bridge when river width is
more than the width given by the Lacey formula. In this case
increase in discharge intensity at the bridge section, causes
bed scour on account of sediment transporting capacity becoming
locally more than sediment supply. Estimation of scour and its
effect on afflux in such rivers are also important design aspects.

Width and depth of alluvial rivers are primarily governed

by discharge. Relationships have been evolved for estimation of
normal dimensions of these parameters. Width and depth formulae
evolved by Lacey are commonly adopted in fixing bridge waterway
and estimating scour depths. Limitations of such formulae and
their implications, however, need to be appreciated to enable
judicious application.

Thus certain topics in open channel flow hydraulics are

of special relevance to bridge engineering which are discussed
below. These comprise velocity distribution and discharge
formulae, afflux due to various reasons and consequent scour,
and formulae for river parameters like width and depth.

52
4.1 VELOCITY DISTRIBUTION AND FLOW FORMULAE

Variation in velocity along a vertical in channel cross

section is known to follow logarithmic distribution, maximum
velocity being at the top and minimum at the bottom. Mean
velocity is obtained at 0.632 depth from top. Average of velocities
measured at 0.2 and 0.8 depths from top also gives the mean
velocity. The velocity distribution along a vertical depends on depth
and relative roughness of bed material. If bottom roughness is
increased, the bottom velocity is decreased and top velocity
correspondingly increased. Distribution curves for different bottom
roughnesses are shown in Fig. 4.1.

River discharge can be determined as A x V where A is

area of flow section and V the mean flow velocity. Sectional area
can be derived knowing cross sectional dimensions. Mean
sectional velocity of flow in a river can be estimated by adopting
any of the following formulae.

(i) R 2 /3 S1 /2 Manning Formula

V=
n
(ii) V = CR 1 / 2S1 / 2 Chezy Formula

fLV 2
(iii) h= Darey Weisbach Formula
8 Rg
V R
(iv) = 6.25 + 5.75log Logarithmic formula
V* Ks

wherein V is the mean velocity in m/s of the cross section, R is

the hydraulic mean depth in m, S is the energy slope which
becomes equal to water surface slope and bed slope under
uniform flow conditions, n the rugosity coefficient, C the Chezy
coefficient, h is the head loss in m, f is the friction factor, L is the
length in m over which head loss h is obtained, g is gravitational
acceleration, V* is friction velocity in m/s equal to (To/ρ)1/2 or
(g RS)1/2 and Ks is the bottom roughness in m. All the above
formulae are of the same form but having different structure and
can be rearranged and expressed as:

53
 37
1.0

H 1000
0.8 K 100
10
0.6
D
H
0.4

0.2

0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
V
V mean
D = TOTAL DEPTH OF HEIGHT.
K = BOTTOM ROUGHNESS.
H = HEIGHT FROM BOTTOM ON A VERTICAL II
V = VELOCITY AT HEIGHT H.
V mean = MEAN VELOCITY OVER A VERTICAL

Fig. 4.1 : Vertical velocity distribution with different

bottom roughnesses

H. F. L.

P1 A1 A2 A3 A4 A5 A6 A7 P
7

P3 P6
A1 A2
R1 = ; R2 = ; P2
P1 2 1 P2 2 1 P5
3 2 3 2 P4
R1 S R2 S
V1 = n1 ; V2 = n2
Q1 = A1 V1 ... Q2 = A2 V2
Q = SQn
Fig. 3.2 : Discharge computation for a river with
composite cross section

54
 1/ 2
V R 1/ 6 C ⎡8⎤
= = 1/ 2 = ⎢ ⎥
V* n g ⎣f ⎦
wherein n, C and f are measures of boundary roughness in
Manning, Chezy and Darey Weisbach formulae.

In case of Manning formula, value of rugosity coefficient

n is given by n=Ks1/6 / 76.5 wherein Ks is median grain roughness
diameter d50 in mm such that 50 percent of the material is finer
than this size. Data of values of n for different sized bed material
and in different types of channels is given in Tables 10.2 and
10.3 of Chapter 10.

River cross section is rarely uniform. It may have one or

more deep channels and also overbank spill area. There can be
considerable difference in n values in channel portion and overbank
area. In such a composite cross section, discharge of each portion
of the section is required to be worked out separately and then
added to arrive at total discharge as indicated in Fig. 4.2.

4.2 AFFLUX IN RIVERS WITH NON SCOURABLE BED

Afflux is defined as rise of flood level of the river upstream

of a bridge as a result of the obstruction to natural flow caused
by the construction of the bridge. It is caused either due to
constriction of waterway by approach banks or on account of
obstruction to flow caused by piers. Afflux results in increase of
velocities in constricted section and at the obstructions. When
riverbed has nonscourable strata or material, the increased
velocities cannot scour the bed and therefore the boundary
remains unchanged. Estimation of afflux under this condition is
relatively simpler. When, however, the riverbed is sandy, it can
scour easily due to increased velocities at the bridge. Estimation
of afflux and scour in this case involves unstable boundary
condition and is, therefore, more involved. Condition of non-
scourable boundary is first considered below.

4.2.1 Afflux due to constriction of waterway

Specific energy Ef in m per m3 of flow is the energy

measured with respect to bottom and hence comprising potential

55
head D and velocity head V2/2g also in m. Thus Ef equals
summation of D and V2/2g. When bridge waterway is constricted
by making the bridge narrower than the channel width, discharge
intensity q increases at the bridge section. Change in q brings
about corresponding changes in D and V but not in Ef. The
interrelationship amongst these parameters is depicted
graphically in Fig. 4.3 reproduced from Reference 4.1.

Following a line of constant specific head Ef in the

subcritical flow region, it is seen that river discharge with a
particular specific head can pass through the bridge with different
discharge intensities q implying different constrictions. Bigger
the constriction, higher the value of q and lesser the depth D.
The velocity head v2/2g correspondingly becomes higher. Such
progressive constriction of bridge waterway and increase in
discharge intensity is, however, possible only up to a limit which
is reached at critical stage. If further constriction is caused, the
available specific energy will be insufficient to pass the discharge
and hence flow will head up on upstream side of the bridge to
force the discharge through. This heading up then becomes
apparent as afflux and its effect is felt as backwater over long
reaches. On the down stream side the energy built up in formation
of afflux is dissipated by generation of hydraulic jump. Conditions
leading to afflux on upstream and jump on the downstream side
of constricted waterway of a bridge are possible when the bed is
nonscourable. When riverbed is sandy, increased velocities in
the constriction develop bed scour. This condition is considered
separately.

In rivers with clayey bed, sufficient bed scour to obviate

afflux may not be formed initially. If constriction is excessive,
significant afflux can result leading to formation of a hydraulic
jump. Under such conditions, foundations of piers and abutments
are required to be designed sufficiently deep to ensure safety.
The safety criterion is to assume that full scour may be obtained
eventually. Full scour required for obviating jump formation is
such which will avoid development of any significant afflux and
consequent supercritical flow in the constricted bridge section.

56
15.240 50
ENERGY
OF
FLOW CURVES
13.716 45

12.192 40

TH
10.668 35

EP
LD
ICA
RIT
2g

FC
2
Energy of flow Ef = D + V

9.144 30

SO
CU
40 LO
0
37
7.620 25 35
5
0
32
5
30
0
28
0
6.096 20 26
0
24
0
22
E

0
LIN

20
0
H

18
PT

0
DE

4.572 15 16
0
14
0
12
0
10
0
90
3.048 10 80
70 Cusecs/Cemecs Cusecs/Cemecs Cusecs/Cemecs
60
50 400 = 11.319 200 = 5.660 60 = 1.698
40 375 = 10.612 180 = 5.094 50 = 1.415
30 350 = 9.904 160 = 4.528 40 = 1.132
1.524 5 20 325 = 9.197 140 = 3.962 30 = 0.849
300 = 8.489 120 = 3.396 20 = 0.566
10 280 = 7.923 100 = 2.830 10 = 0.283
260 = 7.358 90 = 2.547 5 = 0.141
FT.

5
240 = 6.792 80 = 2.264 1 = 0.028
M.

1
220 = 6.225 70 = 1.981
0
0 FT. 5 10 15 20 25 30
Depth (D)
M. 1.524 3.048 4.572 6.096 7.620 9.144
Fig. 4.3 : Relationship between specific energy,
depth of flow and discharge intensity

57
4.2.2 Afflux on account of obstruction to flow

Even if the river constriction on account of approach

banks is not excessive and bridge waterway is wider than that
causing critical discharge intensity, afflux can still be generated
due to obstruction to flow by piers. Separation eddies forming at
such obstructions result in loss of head which shows as afflux
upstream of the bridge. One of the formulae in popular use in
India for estimation of afflux at bridges is due to Molesworth(4.8)
given below.

h = [(V2 / 17.88) + 0.01524] [(A/a)2-1]

wherein h is afflux in m, V is velocity in m/s in

unobstructed stream, A is unobstructed sectional area of the
river in m2 and a is obstructed sectional area of the river in m2.
Other formulae in wide use are due to Negler and Yarnell(4.3)(4.4).
More recently, the Bureau of Public Roads U.S.A., have adopted
procedure for estimation of afflux detailed in Hydraulics of Bridge
Waterway(4.5).

Affluxed waterway has relevance in providing free board

and clearance, norms for which are given in paras 4.4 and 4.5.

4.3 AFFLUX AND SCOUR IN ALLUVIAL RIVERS

In case of alluvial rivers, the river bed comprises

scourable material like sand. Constriction of bridge waterway
causes increase in discharge intensity and increase in velocity
which result in bed scour. Scour in riverbed increases the depth
of flow and reduces velocity and hence the scouring capacity.
When excess of sediment transporting capacity in the constriction
beyond normal capacity in unconstricted reach is neutralised by
progressive bed scour during rising flood of a hydrograph, no
further scour occurs. This depth with full scour approximates
normal scour depth in the same river adjusted for the increased
discharge intensity on account of constriction. With development
of full scour, afflux becomes negligible as witnessed in alluvial
rivers carrying sustained floods. In case of Ganga river at
Mokamah bridge, the afflux observed was only 5 cms for a flood
discharge of 51000m3/s(4.6).

58
 When the river carries a flashy flood, time may be
inadequate to develop full scour, even if the river is alluvial and
bed is scourable. Substantial afflux can then occur. Scour depth
and afflux for a flashy flood can be estimated by making
computations for progressive rise in flood discharge starting with
the lowest stage. Flood hydrograph is converted to a step
diagram, each step pertaining to a small rise in discharge obtained
for a specific duration. Smaller this specific duration, higher will
be the accuracy. For the first rise in discharge, afflux is initially
computed for no scour condition. Scour depth for increased
velocity is then estimated using any of the appropriate sediment
transport functions. Numerous such functions have been given
in Reference 4.7. Time available for scour is the time during
which, that particular discharge stage is experienced. This time
is woven in the method of estimation of the scour depth. With
this scour depth, the afflux is recomputed. By iteration process
the afflux and scour for the first rise in flood discharge is finalised.
Next rise in discharge is then considered. Entire flood hydrograph
is thus covered up to the peak point. Afflux and scour values
associated with the peak stage of a given design hydrograph
provide necessary design data. More detailed procedure is
explained in Reference 4.8.

4.4 FREE BOARD

The free board above designed flood level in case of rivers

and full supply level in case of canals, including the afflux, to the
formation level of the railway embankment or guide bund should
not be less than 1 m. In case where heavy wave action is expected,
the free board should be further increased to allow for the same.

Vertical wave run up depends on bank slope, wave height

and wave period. Wave characteristics such as wave height, length
and period in turn are dependent on wave velocity, straight length
over which it is incident, termed fetch, wind velocity and depth of
water. Knowing this design data, it is possible to estimate wave
run up. Stevenson formula modified by Moliter is in popular use
for estimation of wave height which is given below.

h = 0.0322 (V.F)1/2 + 0.76 – 0.89 F1/4

59
wherein h is wave height in m, F is fetch in km and V is wind
velocity in km per hour.

4.5 CLEARANCE

As applied to the bridge super-structure over water

channels, this is vertical height between the designed flood level
(including afflux) of the stream and a point on the bridge super-
structure where the clearance is required to be measured.

In the case of an arch bridge, the clearance will be the

vertical height between the design flood level (including afflux)
and any point on the soffit of the arch where clearance is required
to be measured.

In the case of road-over or under bridge, it is the shortest

distance between the moving dimensions for railway or road
vehicles and the superstructure.

In case of bridges with rectangular openings, the

minimum clearances normally provided are as in Table 4.1.

Table 4.1

Discharge Vertical Limits upto which relaxable

(Cumecs) Clearance in special circumstances
(mm) (mm)

Less than 3 600 300

3 – 30 600 300 – 400 (Prorata)
31 – 300 600 –1200 400 – 1200 (Prorata)
301 – 3000 1500 No relaxation is permissible
Above 3000 1800 No relaxation is permissible

60
 In case of arch bridges, minimum clearances measured
to the crow of the arch are indicated in Table 4.2.

Table 4.2

Span of Arch (m) Clearance from crown of arch

Less than 4 Rise or 1200 mm whichever is more

4–7 2/3 Rise or 1500 mm whichever is more
7 – 20 2/3 Rise or 1800 mm whichever is more
Above 20 2/3 Rise

In aggrading rivers where the bed has a tendency to rise

the clearances provided should be suitably augmented. Bigger
clearance may be required when river navigation is to be catered
for. Possibility of floating trees being brought down from upper
catchment needs to be considered and adequacy of clearance
checked on this accord.

4.6 LACEY FORMULAE

Lacey originally evolved width and depth formulae for

stable alluvial canals. These were the following. (4.9)

P = 4.836 Q1/2

R = 0.473 ()
Q
f
1/3

wherein P is wetted perimeter in m, Q is discharge in

m3/s, R is hydraulic mean radius in m and f is the silt factor
which is given as 1.76 m1/2, m being weighted mean diameter of
bed material in mm. Working out f value from frequency diagram”
of river bed material is indicated under para 6.8.1 (i) of chapter 6.

On the basis of data of a few rivers, Lacey showed that

his P:Q relationship is valid not only for canals but also for alluvial
rivers. Since river sections are much wider than canals, P and R
can be replaced by W and D, namely width and depth respectively.

61
 Bridge waterway is normally designed using Lacey
relationship P = 1.811 CQ1/2 allowing some deviation in the value
of C from 2.5 to 3.5 according to local conditions. The depth of
relationship D = 0.473 (Q/f)1/3 is employed for calculation of scour
depth at bridge piers or along guide bunds adopting multiplying
scour factors given in I.R. Substructure Code as given in Table
4.3.

Table 4.3

S.No. Nature of the river Depth of Scour

1 In a straight reach 1.25 D
2 At a moderate bend condition 1.5 D
e.g. along apron of guide bund
3 At a severe bend 1.75 D
4 At a right angle bend or at 2.0 D
nose of piers
5 In severe swirls e.g. against 2.5 to 2.75 D
mole head of guide bund

An important aspect of the two basic Lacey formulae is

that they are derived and applicable for near bankful stage along
the river but are invalid and should not be used to find variation in
either W or D at a place at different discharges or flood stages.

Secondly, the width formula was shown by Lacey to be

applicable to rivers on basis of only 7 observations. Rivers of
braided pattern are, however; known to be wider than Lacey width.
For instance Brahmaputra and Sone rivers have ratios of actual
width to Lacey width of as much as 12.41 and 7.8 times
respectively. On the other hand in case of incised rivers the actual
width can be shorter than Lacey width. In Tapi river, the width is
0.44 times that given by Lacey formula. Generally the Lacey
width formula gives a better fit for alluvial rivers in flood plains. A
few typical examples of variation in width ratio in different types
of rivers are given in Table 4.4.

62
 Table 4.4

Comparison of actual widths of rivers with those given by

the Lacey Formula

S. River Bankful Actual Lacey Actual Remarks

No. discharge River Width ‘W’/Lacey
m 3/S Width (m) (m) ‘W’
1 Jhelum, 394 103 97 1.06 Flood Plain
J&K, India Meandering
2 Bhagirathi, 1670 218 198 1.10 -do-
West Bengal
India
3 Ramganga 6600 345 394 0.88 -do
U.P. India
4 Indus of 7050 985 1020 0.97 --do
Hajipur
Pakistan
5 Mississippi 42400 1380 1000 1.38 --do-
at Vieksberg,
USA
6 Tapi, Gujarat 17000 274 630 0.44 Incised
India
Flood Plain
7 Kosi, Bihar 7050 6150 406 15.06
Braided
India
8 Brahmaputra 24700 9450 760 12.41 Flood Plain
Assam India Braided
9 Sone, Bihar 14100 4500 575 7.79 Flood Plain
India Braided
This limitation of Lacey formula needs to be borne in
mind while using it for estimation of bridge waterways.

Deviation in respect of depth parameter is less than that

of width. Better fit to field data is possible if Lacey depth formula
expressed in terms of ‘q’ the discharge intensity in m3/s per m
width instead of ‘Q’ the total sectional discharge in m3/s, giving
depth ‘D’ in m as

63
 D = 1.34 (q2/f)1/3

‘f’ being the Lacey silt factor. This formula can be reduced to

D = 1.12 Q0.67 W-0.67 m-0.17

where ‘W’ is water surface width in m and ‘m’ the size of bed
material in mm. The semi-theoretical formula derived by Laursen
and Latishenkov curves in vogue in the U.S.S.R. have the same
form of equation as the Lacey D,q, f formula.

The formula D,Q,f and D, q, f become identical when

width and average depth of a river section are equal to those
given by Lacey Formulae. When width, however, is different from
that given by Lacey formula, D, q, f formula become applicable in
preference to D, Q, f formula.

Important Conversions

In F.P.S. In Mks

K1S/ 6 K1S/ 6
n= (KS in ft) n= (KS in mm)
29.3 76.5

⎡ V2 ⎤ ⎡⎛ A ⎞2 ⎤ ⎡ V2 ⎤ ⎡⎛ A ⎞2 ⎤
h =⎢ + 0.05⎥X⎢⎜ ⎟ − 1⎥ h = ⎢ + 0.01524 ⎥X⎢⎜ ⎟ − 1⎥
⎣ 58.6 ⎦ ⎣⎝ a ⎠ ⎦ ⎣17.88 ⎦ ⎣⎝ a ⎠ ⎦
1/2 1/4
h = 0.17 (V . F) + 2.5 - (F) h = 0.0322 (V . F)1/2 + 0.76 – 89(F)1/4
1/2
P = 2.67 Q P = 4.836 Q1/2
1/3
R = 0.473 (Q/f) R = 0.473 (Q/f)1/3
2 1/3
D = 0.9(q /f) D = 1.34(q2/f)1/3

64
REFERENCES

4.1 Khosla A. N., Bose N. K., Taylor E. M. ‘Design of weirs

on permeable foundations’. Central Board of Irigation and
Power, India, Publication No. 12, 1936.
4.2 ‘Indian Railway Standard - Code of practice for design of
substructures of bridges’, revised 1985, Government of
India, Delhi.
4.3 ‘Manual on River Behaviour, Control and Training’, Central
Board of Irrigation and Power, Government of India, New
Delhi, Publication No. 60', Revised 1971.
4.4 Yen Te Chow ‘Open Channel Hydraulics’; Mac Graw-Hill
Book Company 1959.
4.5 Bradley J.N. “Hydraulics of bridge Waterways’, Hydraulic
design series No.1, Bureau of Public Roads, U.S.A.,
Second Edition, 1970.
4.6 Joglekar D.V., Discussion in Transactions of ‘Scour at
Bridge Crossing by Laursen E.M., Journal of Hydraulics
Division, American Society of Civil Engineers, February,
1960.
4.7 American Society of Civil Engineers, ‘Sedimentation
Engineering’, Manuals and Reports on Engineering
Practice No. 54, 1975.
4.8 Gole C. V., Chitale S.V., Rajgoplan K. S. ‘Water-Way
for bridges across alluvial stream, International
Association for Hydraulic Research, 1975 Congress, Sao
Paulo, Brazil.
4.9 Lacey G., ‘Stable Channels in alluvium,’ Paper No. 4736,
Minutes of proceedings of Institution of Civil Engineers,
Vol. 229,1930.
4.10. Inglis C.C., ‘The Behaviour and Control of Rivers and
Canals’, Central Water and Power Reserach Station,
Poona, India, 1949.
4.11 Izbash and Khaldre KH Yu, ‘Hydraulics of River Channel
Closure’, Translated from the Russian by Chairns G.L.,
London, Butter worths, 1970.
4.12 U.S. Army Corps of Engineers,. Waterways Experiment
Station, ‘Hydraulic Design Criteria -Sheet 712-1’,
Vicksberg, Mississippi, U.S.A.
4.13 Hunter House, “Engineering Hydraulics’, John Wiley
and Sons, 1950, page 126 .

65
 Chapter 5

TYPES OF TRAINING AND PROTECTION

WORKS

5.1 BRIDGE LOCATION AND ALIGNMENT

A bridge is normally located where the river section has

minimum width so that the structure becomes shorter and
economical. As far as possible it is aligned normal to the river. If
satisfying these requirements involves a long detour, some
compromise is made in siting a crossing. Care is also taken to
avoid, if possible, sites where frequent changes occur in river
course, tendency for aggradation or degradation is manifest and
problems of bank erosion or difficult foundation conditions are
required to be faced.

Approach banks in case of constricted bridges are

generally aligned in line with the bridge axis and normal to the
guide bunds. Approaches deviating towards downstream direction
add to their safety from river loops approaching dangerously close
while deviation in upstream direction exposes them more to river
attack.

Location of the bridge with respect to Khadir width also

determines the length and position of approach bank exposed to
river attack. When bridge is constructed near one of the ends of
the Khadir width and when the khadir edge consists of stiff material
resisting erosion, cost of one guide bund can be saved. Longer
length of the approach at other end however gets exposed and
becomes vulnerable to river attack in this case. On the other
hand central location of the bridge reduces length of approach
open to river attack, minimises possibility of extreme obliquity of
approach of the river towards the bridge but requires provision of
guide bunds on both the flanks.

In general, bridge location and alignment are fixed to

ensure normal approach of the river, equitable distribution of flow
across bridge section and minimum possibility of river attack on
approach banks.

66
5.2 BRIDGE WATERWAY

Larger bridge opening involves more cost. It permits

haphazard sediment deposition and formation of islands. Obliquity
and concentration of flood flow can then develop any where on
bridge section with consequent deep scour and high velocity. In
India, the practice evolved in case of alluvial rivers is, therefore,
to constrict the river waterway at bridges. Thus the main
advantages of constricted waterway are the resulting economy
and better hydraulic performance of the bridge. The first bridge to
be constricted was on Chenab river at Shershah in 1888, (5.1) now
in Pakistan, shown in Fig. 5.1.

Benefits accruing from constriction were so convincing

that constriction of bridges on alluvial rivers has since become a
standard practice.

Extent of permissible constriction was determined by

Spring and Gales(5.2) on basis of river data of maximum flood
depths and velocities. Lacey on the other hand verified the validity
of his relationship between wetted perimeter P and discharge Q
originally evolved for alluvial canals for application to rivers. (5.3).
Parameter P can be replaced by width W in case of wide rivers
giving

W = 4.836 Q1/2

wherein W is in m and Q in m3/s. His finding was that this formula

fits the river data well and hence can be adopted for determining
constricted waterway of a bridge. Subsequently Sethi (5.4) and
Khosla Committee(5.5) accepted Lacey formula as basis for design
of waterway with suitable deviations. The current practice is in
accordance with the recommendations of the Khosla Committee.

Excessive constriction of bridge waterway results in

afflux. In case of alluvial rivers, afflux causes bed scour which
nearly compensates the waterway carrying capacity lost in
constriction and hence afflux becomes negligible. When riverbed
is non-scourable or can only be partially scoured, afflux cannot
get obliterated and constriction of Lacey width may not be
justifiable.

67
 CHENAB RIVER
AT SHER SHAH

IN 1890 - 91 AFTER BRIDGING AND TRAINING

RAILWAY
CA
NA
L
AD
RO

Fig. 5.1 : Railway bridge on Chenab river at Sher Shah

68
5.3 GUIDE BUNDS

Constriction of waterway at bridges is effected by

extending the approach banks to cover spill area or, in addition,
part of active channel. Such encroachment causes obstruction
to flow. Guide bunds are then required to be provided for guiding
flow smoothly through the bridge.

The course of main channel may change on account of

progressive movement of meanders or formation and movement
of islands in braided rivers. In the process, attack may develop
on bridge approaches. Under such conditions it becomes
obligatory to provide guide bunds in order to pass the river
discharge axially with as uniform flow distribution through the
bridge as possible. Guide bunds also prevent the river cutting
into the bridge approaches, causing breaches and forming deep
scour at abutments. These objectives are achieved by giving guide
bunds suitable shape, providing adequate length, by adding curved
heads and by giving suitable protection to exposed faces against
river attack.

5.3.1 Shape of guide bunds

Alignment of guide bunds converging on the upstream

was considered by Bell, Spring and Gales to allow for the waterway
obstructed by piers on bridge line as indicated in Fig. 5.2 (B).

With such a shape, flow may separate after passing

round upstream curved heads leading to shoal formation on flanks
near the bridge axis rendering end bays inactive. Divergent guide
bunds were conceived by Haigh and Spring to mitigate tendency
of separation. The distance between the worst loop of the river
and the approach bank however gets reduced and shoal formation
in the centre in the wide section some time occurs as in Fig.
5.2(A). Symmetrical and parallel guide bunds were proposed by
Spring and Sethi as depicted in of Fig. 5.2(C). When provided
with upstream curved heads of sufficiently big radius, this shape
of guide bunds is found to prevent separation of flow along flanks
and help in effecting equitable distribution of discharge across
the bridge section. It is also suitable under condition of variable
approach. Elliptic guide bunds were advocated by Sharma et.
al.(5.6) of the Uttar Pradesh Irrigation Research Institute (U.P.I.R.I.)

69
 RIVER CURRENT

HIGH BANK
SHOAL

WORST POSSIBLE
EMBAYMENT

AXIS

(A) DIVERGENT UPSTREAM

RIVER CURRENT

HIGH BANK
WORST POSSIBLE
EMBAYMENT
SHOAL

AXIS

(B) CONVERGENT UPSTREAM

RIVER CURRENT
HIGH BANK

WORST POSSIBLE
EMBAYMENT

AXIS

(C) PARALLEL

Fig. 5.2 : Guide Bunds of Differentshapes

70
in case of wide and shallow rivers to induce the flow to hug the
guide bunds better without separation all along their lengths as
shown in Fig. 5.1. Considering all the above shapes, in case of
alluvial rivers with sandy bed and meandering pattern, elliptical
shape appears preferable to minimise obliquity and separation
of flow. Straight parallel guide bunds with composite curve for
upstream heads have also been found to give optimum design
as in case of Mokameh bridge on Ganga river. In case of rivers in
submontane region, in rivers with bed material coarser than sand
size and in rivers with braided channel pattern, experience is
that nonsymmetrical guide bunds with different length· and shape
on left and right banks may be warranted on account of local
conditions and may be found to result in superior hydraulic
performance.

5.3.2 Length of guide bunds

Length of a guide bund has to be such that the flow

approaching even from behind the guide bund would pass through
the bridge with minimum obliquity ensuring optimum utilisation
of waterway. Length has also to be adequate to keep worst
embayment of the river at the back of the guide bund away from
the approach bank at a safe distance.

For coaxing the river to flow axially through the bridge

lengths of guide bunds considered essential by Spring, Gales
and Sethi are given in Table 5.1. Recommendations of Gales are
for limiting obliquity of flow through the bridge from 300 to 340.

In addition, bend flow analogy is relevant in this case.

The analogy of flow, downstream of a curved channel indicated in
Fig. 5.3 suggests that flow distribution within guide bunds would
become more uniform if radius of curved head is increased rather
than increasing the length of the guide bunds as was exemplified
by the left guide bund of the Kosi barrage.

71
 Fig. 5.3 : Velocity distribution in channelsof different
curvatures

Table 5.1 : Length of Guide Bund

Portion According According According to Remarks

to to Gales Sethi
Spring
1 1 1
Up Stream 1 10 L 1 4 L 9/10 to 1 10 L (i) Length of guide
Length bund is straight
for discharge of in flood plains length between
7086 to 21254 and longer as bridge alignment
m 3 /s required for and parallel line
rivers in joining apex of
submontane upstream curved
region. heads.

Down Stream 1
/10 L to 1/5 L ¼ L for discharge 1
/5 to ¼ L for (ii) ‘L’ is length of
Length of 7086 to 21254 rivers in flood bridge between
m 3 /s plains and abutments.
longer as
3/8 L for equired for
discharge of rivers in
42507 to 70847 submontane
m 3 /s region.

72
 Procedure advocated by Spring, Gales and U.P.I.R.I. for
determining length of guide bund to keep away the worst bend
forming in the river was to first ascertain dimensions of such a
bend obtaining in the river when a cut off occurs. Ratio of bend to
chord at this stage in lower Ganga river was found by Gales to
be 1.75. When such acute bends are not found in available
surveys, radius of worst loop can be estimated on basis of average
radius of existing bends. For each bend, the meander length ML
along valley and meander width MB across valley is measured.
Radius of curvature R can then be roughly worked out assuming
the meander shape to have formed of circular arc and using the
relationship.

(0.25 ML)2 = (MB – W) [R – 0.25 (MB – W)]

wherein W is the average width of main channel during floods.

Mean of the R values of all bends is found which then permits
estimation of radius of worst bend using the ratio of R for average
bends to R of worst bend varying inversely as discharge from 2.5
to 1.7 for discharges between 2000 to 9000 m3/s as found at the
U.P.I.R.I(5.7). Garg and others from the U.P.I.R.I. proposed fitting
such a curve behind guide bund when Khadir width is small and
double bend when Khadir width is big. The procedure of fitting in
worst bend at the back of the upstream of guide head bund is
illustrated in Fig. 5.4.

1200 WORST ANTICIPATED

LOOP OF RIVER
RI
VE

120
0 885 m.
R

885 m.
AXIS OF
STRUCTURE

60 m.
1250 m.

Fig. 5.4 : Fitting of worst bend at curved head

of a guide bund

73
 Length of guide bund determined on basis of the above
two considerations of axial flow through the bridge and keeping
the worst bend away at a safe distance may still be found to be
inadequate to protect the long approaches in rivers having wide
khadirs. Such a situation may be met with more frequently in
case of braided rivers. Instead of elongating the guide bunds
further, it would be advisable in such rivers to provide additional
training works such as spurs, revetments etc. for protection of
approach banks.

5.3.3 Curved heads

Several considerations govern the radius of curved head.

More the radius and sweep, the cost becomes more. Minimum
curvature necessary for running railway line to carry stone and
construction materials is another important consideration. For
broad gauge, curve having radius of 250 m is found to be suitable.
Minimum curvature necessary to prevent separation of flow from
the curved head has also to be determined and provided.
Otherwise separation eddies would form deep and dangerous
scour along the guide bunds.

According to Spring, water with a velocity of 2.4 to 3.0

m/s obtaining during floods can easily follow along a curve of 183
to 244 m radius. For lesser velocities radius can be reduced
according to Table 5.2 as suggested by him.

Garg and other from the U.P.I.R.I. evolved, on the basis

of model studies pertaining to 16 different structures, the following
relationship for radius of curved head Rc (5.7) in m,

Rc = 0.45 Pw

wherein Pw is the width in m given by Lacey Formula.

In case of Mokamah bridge over Ganga river, composite

curve was evolved on basis of model studies. For the upstream
head, 20 and 30 curves were provided upto 900 sweep and 5o curve
between 90o and 1200 where velocities were relatively lower as
shown in Fig. 5.5 (Degree of a curve is obtained by dividing 1750
by its radius in m).

74
 Table 5.2

Radius of Curved Head According to Spring

Probable Radius of upstream

Sand maximum curve head of guide
bund (m) for Average Remarks
Size abnormal scour
below bed level fall of river in cms/km
4.66 9.32 13.98 18.64 27.96
Very Under 6.10 m 61 76 91 107 122
Coarse Over 6.10 m 76 95 114 134 152 Radius of down-
stream curved
Coarse Under 9.15 m 91 110 130 149 168 head should be
Over 9.15 m 107 131 156 180 204 half of that of up-
stream curved
Medium Under 12.20 m 122 130 168 191 213 head subject to a
Over 12.20 m 137 168 198 229 259 minimum radius
required for main-
Fine Under 15.25 m 152 180 206 232 259 tenance trains to
Over 15.25 m 183 221 252 282 311 be shunted
Very Fine Under 18.30 m 183 213 244 274 305
Over 18.30 m 244 274 305 335 306

Recommendations in respect of upstream and

downstream curved heads and their angle of sweep made by
Spring, Gales and Sethi are summarised in Table 5.3.

Direct comparison of radii of curved head obtained by

following different practices is not possible. With certain
assumptions and approximations, it was however possible to
derive the following three relationships. (5.8)

Rc = 71.80 Q0.129 on basis of Spring’s Table

Rc = 0.935 Q0.575 on basis of Gales and Sethi recommendations.
Rc = 2.4 Q0.5 on basis of recommendation of Garg et al of
U.P.I.R.I

75
Fig. 5.5 : Composite curve for guide bund head of
mokame bridge on Ganga river

76
 Table 5.3

Radius of Curvature and Angle of Sweep According to

Spring, Gales and Sethi

Item According to According According

Spring to Gales to Sethi

Radius R in m of R = 183 to 244 m R = 249.6 m i.e. R = 249.6

upstream for 7 degree curve for to 582.2 m
curved V = 2.4 to 3.0 m/s, Q = 7086 to with
head Radius depends 21254 m3/s and composite
on slope, scour R = 582.2 m i.e. curvature
depth and size of 3 degree curve
bed material as in for Q = 42507 to
Table 5.2 70847 m3/s
maximum radius
being 366 m.

Radius R in m (i) 61 to 91 m R = 249.6 m

of downstream (ii) Half of i.e. 7
curved head upstream head degree
(iii) Minimum curve
required for
shunting
maintenance
trains

Upstream head 120o to 140o 120o to 140o 120o to

angle of sweep 140o

Downstream 60o 60o

head angle of (and not 90o)
sweep

77
 Radius of curvature arrived at using these relationships
for various discharges are given in Table 5.4.

Table 5.4 : Radius of Upstream Curved Head of Guide

Bund According to Various Investigators

Radius of Curvature Rc in
Discharge Q
Metres according to
m3/s
Spring Gales Garg

283 148.73 24.02 40.37

1416 183.07 60.63 90.31
2832 200.19 90.32 127.72
5564 218.41 133.17 179.02
7080 225.31 152.96 201.94
14160 246.38 227.86 285.59
21240 259.61 287.69 349.77
28321 269.43 339.45 403.89
42481 283.90 428.57 494.66
56641 294.63 505.66 571.18
70802 303.24 574.89 638.61

For a discharge of 14160 m3/s, Spring recommended

radius of about 247m. Gales and Garg also give comparable
values of 228 and 284m respectively. However, for small
discharges of the order of 283 m3/s Spring’s value of Rc is more
than six times Gale’s while for high discharge of 70802 m3/s,
Spring’s value is almost half of that given by Gales and Garg. At
the present stage confirmatory and conclusive evidence for
supporting anyone of the above practice is not available. It is,
therefore, advisable to follow Spring’s Table for discharge upto
14160 m3/s and Garg’s recommendation for higher discharges.

78
5.3.4 Protection for side slopes

Guide bund is normally constructed with sand core with

side slopes of 2: 1 and provided with stone protection on river
side which is continued along the back of curved heads., Clay
should not be used for constructing guide bunds since it is subject
to heavy settlement and clays soluble in water are liable to be
sucked out by the current. On the backside of the shank turfing
on clay blanket is given. The material for core should be obtained
from the riverside and not from the back side. The weight of stone
according to Spring should be between 25 and 50 kg. He observed
that such stones, called one man stones, can withstand velocities
up to 5.5 m/s. Stones should be angular and not rounded and
should be durable.

Thickness of stone protection recommended by Spring

is given in Table 5.5.

Table 5.5

Size of Thickness of stone protection (cm) for

riverbed average fall of the river (cm/km)
material 5 14 19 28 37
Very coarse 40 47.5 55 62.5 70
Coarse 55 62.5 70 77.5 85
Medium 70 77.5 85 92.5 100
Fine 85 92.5 100 107.5 115
Very Fine 100 107.5 115 122.5 130

He suggested that thickness can be reduced by 150 to

225 mm by using quarry refuse or burnt bricks as a filter. The
thickness was recommended to be increased by 25 percent at
head to take care of severe attack and 25 percent all over when
pitching is dropped through deep water below low water level.

Thickness of pitching and soiling for guide bund slope

recommended by Gales is given in Table. 5.6. Soling formed by
stone ballast to act as filter was proposed by him.

79
 Table 5.6
Thickness of Stone and Soiling for Slope for Guide
Bunds According to Gales

River Discharge (m3/sec)

7086 to 21254 21254 to 42507 42507 to 70847
Item U/S Body U/S Body U/S Body Remarks
head and head and head and
D/S D/S D/S
head head head
Thickness of Pitching
pitching stone 105 105 105 105 105 105 stone to be
(cm) hand set.

Thickness of Ballast to be
soiling ballast 17.5 17.5 20.0 20.0 22.5 22.5 broken to
(cm) pass 6.25
cm ring

Total thickness 122.5 122.5 125 125 127.5 127.5

of slope
covering (cm)

Inglis(5.9) proposed that thickness of pitching be made

dependent on discharge such that

T = 0.06 Q1/3

wherein T is thickness in m and Q the discharge in m3/s.

Sethi suggested that thickness should be varied in

accordance with discharge as well as sand size following the
relationship.

()
1/3
Q
T=K F

wherein K is the coefficient varying inversely with the discharge.

This variation was given graphically as shown in Fig. 5.6 and in
tabular form as in Table 5.7.

80
81
Fig. 5.6 : Thickness of pitching recommended by Sethi
 Table 5.7 : Thickness of stone pitching for shank according to Sethi

S.No. Discharge Soiling Thickness of Total

(m3/sec) thickness Pitching Stone Thickness Remarks
(cm) (cm) ‘T’ (cm)

1 2832 15 75 90 Values in this Table were worked out using the

relationship T=K(Q/f)1/3 and adopting values of
K for various values of Q from figure 5(b) given
2 7079 15 82.5 97.5 by Sethi in Reference 4.

3 Thickness of stone pitching in underwater slope

11327 15 90 105
should be 1.33 T for shank and downstream

82
head.
4 14159 15 97.5 112.5
Use of 45 to 60 cm thick cement concrete slabs
on slopes supported by a suitable toe may be
5 28317 15 105 120
considered in place of pitching stone where
costs of the two arrangements compare
6 42475 22.5 105 127.5 favourably.
Thickness of stone pitching for U/S head above
7 56634 22.5 112.5 135 low water level should be same as that of shank.
For underwater slope, the thickness for U/S
8 70790 22.5 120 142.5 head should be 1.5 T.
 Thickness for underwater slope was stipulated to be’
1.5 T for head and 1.33 T for the body and tail.

Considering practices advocated by different authors,

one man stones of 30 cm to 40 cm size weighing 40 to 70 kg are
considered adequate to withstand velocities up to 3.5 m/s which
are commonly met with in alluvial rivers in flood plains. The
thickness of pitching is, however, varied to meet different
situations. Spring proposed variation in thickness according to
river slope and size of bed material, according to possibility of
providing filter, according to difficulties in achieving proper standard
of work as for underwater placement and according to exposure
to attack as at curved head. Gales and Inglis proposed variation
in thickness according to discharge while Sethi considered size
of bed material as additional relevant parameter.

Apart from the above Indian practice, theoretical

approach to the problem of design of pitching is also possible.
Stone can get dislodged due to hydrodynamic drag and lift caused
by velocity of flow. For preventing dislodgement either by sliding
or by overturning, the size of stone is required to be varied as the
square of the velocity. Size or weight necessary for stability of
an isolated stone is much more than in case of stone surrounded
by others when laid to form a layer. Incidence of direct attack
caused by oblique flow, formation of eddies, higher level of
turbulence are some other factors requiring heavier stones for
stability against a current of given velocity. For the same velocity,
rounded stones, stones with higher vertical dimension, stones
with smaller specific gravity and stones on steeper side slope
are less stable. For stability of stone pitching against current
velocity on a side slope, the slope angle has to be flatter than
the angle of repose of pitching material. On account of such
numerous factors governing stability of stone, value of C in various
relationships of the form

Velocity V = C1 (diameter D)1/2

can be different. Since volume and weight of a spherical body

are interrelated, this relationship can as well be expressed in
terms of weight, Weight W = C2 (velocity V)6

wherein W is the weight of a sphere having the same volume as

83
that of the stone and C1 and C2 are coefficients. Some such
relationships are presented in Table 5.8.

Normally the side slope of guide banks is kept at 2

horizontal to 1 vertical. For this side slope, relationship
recommended by the Indian Standards Institution in IS 8408-
1976 for protection of side slope is
W = 0.02 V6
which is shown plotted in Fig, 5.7 reproduced from Fig. 6 in IS :
8408-1976.

According to railway practice one man stone used for

protection works has a weight of about 55 kg (120 Ibs) and can
withstand velocities up to 3.5 m/s (11.5 ft/s) or so. IS: 8408:1976
also states that one man stone weighing 40 to 70 kg (giving
mean weight of 55 kg.) can withstand velocities upto 3.5 m/s.
For W of 55 kg and Y of 3.5 m/s, the relationship of the form
W = CV6 gives as the value of C as 0.03.

In IS: 8408-1976, thickness of pitching stone given in

clause 5.5 is equal to size of stone but not less than 0.2 m. In
railway practice, the thickness of pitching stone based on
recommendations of Spring, Galses, Sethi, Inglis, etc. is much
more in vogue. If thickness is to be reduced to conform to ISI
specifications, the size or weight has to be bigger. In that case
the coefficient should be even more than 0.03, say equal to 0.04,
giving
W = 0.04 V6
in preference to relationship
W = 0.02 V6
recommended in IS : 8408-1976 for determining size and weight
of stone on the slope of the guide bunds.

Regarding apron protection two limiting curves are given

in Figure 7 of IS: 8408-1976 which is reproduced as Fig. 5.8,
depicting relationships

W = 0.0161 V6 for surrounded stone, and

W = 0.1003 V6 for isolated stone.

84
 Table 5.8 : Relationships for Size and Weight of Stone
S.No. Relationship Relationship for Sponsoring Source Remarks
(1) for Size weight Agency (5) (6)
(2) (3) (4)
1. Vb = 5.95D1/2 W = 0.031 Vb6 U.S. Bureau of Adopted from the report of Sub- Relationship is for 2:1 side slope. T.V.A. adopted
(Vb is velocity Public Roads Committee on slope protection similar procedure vide reference 5.11. Depend-
in A.S.C.E., proceedings, ing on the severity of attack Vb is increased up
against stone) June, 1948 to 100 percent

W = 0.254 Vm6 U.S. Army Corps Relationship originally given in Relationship is for 2:1 side slope and derived
2. Vm = 4.2 D1/2 Ref. 5.12 is T=0.04 (rs-r) D50. from relationship in terms of T given for horizon-
of Engineers This was converted by author tal bottom.

85
(Vm is local
average velocity) as Vm= 4.9 D1/2 and then side
slope effect allowed for as sug-
gested in Reference 5.12 which
gave Vm= 4.2 D1/2 for 2:1 side
slope
The curve is given in Hydrau- Meant for protection of channel bottom and side
3. Vm = 5.65 D1/2 W = 0.043 Vm6 United States lic Design Criteria of U.S. wa- slopes down-stream of stilling basins and for
Bureau of ter-ways experiment station, in rock sizes for river closures. Vm in preference
Ref. 5.14, sheet 712-1.It ap- to Vb is recommended to be adopted to account
Reclamation proximates Isbash Curve with for indeterminate flow factors like obliquity of
curve. E = 0.86 attack, eddy action, etc. It nearly corres-ponds
to curve of Bureau of Public Roads with 1:1
side slope.
 S.No. Relationship Relationship for Sponsoring Source Remarks
(1) for Size weight Agency (5) (6)
(2) (3) (4)
4. For SS 1:1 For SS 1:1 California Public IS : 8408-1976 and Draft of ‘Cri- The plots in figure 6 of IS 8408-1976 are based
Vm = 5.00 D1/2 W = 0.089 Vm6 Works Department. teria for Design of Guide Banks on the relationship
0.1136V6SS.Z
for Alluvial Streams’ prepared W=
For SS 2:1 For SS 2:1 by BDC.68 of I.S.I. (SS --1)3
Vm = 6.37 D1/2 W = 0.021 Vm6 wherein Z is Cosec3 (70O-a), Ss is specific
gravity of stone, a is angle in degrees of side
For SS 3:1 For SS 3:1 slope. This relationship was adopted from Cali-
fornia Public Works Department and shown
Vm = 6.81 D1/2 W = 0.014 Vm6 plotted in Fig.6 without giving relationship. They
are recommended by I.S.I for slope protection.

86
CPW Department recommends Vm to be in-
creased by 33% for very violent attack and
reduced by 33% for practically no attack.
 S.No. Relationship Relationship for Sponsoring Source Remarks
(1) for Size weight Agency (5) (6)
(2) (3) (4)
5. Vm = 6.65 D1/2 for W = 0.0161 Vm6 ISI IS 8408-1976 and Draft of ‘Cri- ISI has recommended these relationships for
surrounded stone. for surrounded teria for design of Guide Banks apron protection. These two relationships are
Vm = 4.9 D1/2 for stone, for Alluvial Streams’ prepared shown plotted in Fig. 7 without giving relations.
isolated stone W = 0.1003 Vm6 by BDC 68 of ISI. They are supported by following relationships.
for isolated stone. V = 6.5 D1/2 - Garde, V = 6.8 D1/2 - Isbash for
For 40 kg stone For 40 kg stone surrounded stone and
6. and 3.5 m/s and V = 4.9 D1/2 - Isbash, V = 4.9 D1/2 - Berry,
velocity 3.5 m/s velocity
Vm = 6.32 D1/2 V = 4.8 D1/2 - Mavis Laushy for isolated
W = 0.0218 V6 stone.

87
For 70 kg stone
and 3.5 m/s For 70 kg stones
velocity 3.5 m/s velocity 40 to 70 kg stone of 0.3 to 0.4 m diameter is
Vm = 5.76 D1/2 Indian Practice IS 8408-1976
said to withstand velocities upto 3.5 m/s
Vm = 4.9 D1/2 W = 0.0381 V6

For 55 kg stone For 55 kg stone

and 3.5 m/s and
velocity 3.5 m/s velocity
Vm = 5.99 D1/2 W = 0.0300 V6
 74
3
4
2 10
BANK SLOPE
1:1
1
BANK SLOPE 103
0.8
2:1
STONE DIAMETER m.
0.6

WEIGHT OF STONE Kg.

0.4 102
BANK
SLOPE

(SS = 2.65)
3:1
0.2 10

0.1
1.0
0.08
0.06
-1
0.04 10
0.03
0.7 1 2 4 6 8 10
VELOCITY m/s

Fig. 5.7 : Size pitching stone vs velocity

3
2 104

1
0.8 103
S

0.6
NE
TO

0.4 102
NE
STONE DIAMETER m.

WEIGHT OF STONE Kg.

DS

TO
TE

DS
LA

(SS = 2.65)

0.2
TE
SO

10
LA
RI

SO
FO

RI

0.1
1.0
FO

0.08
-0.06
-1
-0.04 10
-0.03
0.7 1 2 4 6 8 10
VELOCITY m/s

Fig. 5.8 : Size of apron stone vs velocity

88
 The weight of stone when isolated is thus nearly six
times the weight of surrounded stone. Instead of this extreme
factor of 6, if factor of 3 is used, the relationship of apron stone
worked out from relationship of slope stone would be

W = 3 X 0.04 V6 = 0.12 V6

or nearly W = 0.1 V6

or V = 4.9 D1/2

or D = 0.042 V2

This is also the relationship for isolated stone given by

the I.S.I., and hence appears acceptable.

The size and weight so determined can be then increased

arbitrarily to account for direct impingement due to sharp change
in approach direction, for increased severity of attack, degrees
of turbulence, incidence of waves etc. The thickness of pitching
is generally kept equal to and limited to twice the size of stone
and for preventing sucking of underlying material of bank section
through interstices, proper filter is provided in between.

Frequently part of the guide bund slope is required to be

pitched under submerged condition below low water level when
satisfactory provision of filter becomes infeasible. Control in
placement of 1-2 layers of pitching stone under water also cannot
be rigorous. Hence theoretical approach normally becomes
difficult to follow for under water work in absence of special
materials and equipment. The method can, however, be used
when proper standard of work can be achieved, the entire or
most of the slope pitching being in dry. If quality control is difficult,
it is considered advisable to follow the Indian Practice mentioned
before, of providing the thickness of pitching arbitrarily much more
than 2-3 layers, following recommendations of Spring, Gales,
Inglis etc., since it has withstood the test of time at bridges on
all major river systems.

89
5.3.5 Apron protection

5.3.5.1 Scour depth

Apron is provided beyond toe of slope of guide bund so

that when bed scour occurs, the scoured face will be protected
by launching of apron stone. For, arriving at the quantity of apron
stone probable depth of scour needs to be estimated. Lacey
formula,

D = 0.473 (Q/f)1/2

wherein D is depth in m below H. F. L., Q is discharge in m3/s

and f is silt factor is used in this connection to work out flow
depth and this depth is increased by using multiplying factors
given in Table 4.3 to arrive at scour depth below H.F.L. Silt factor
f is given as 1.76 m½ m being weighted mean diameter of bed
material in mm.

5.3.5.2 Scour factors

According to Spring. a fair basis to estimate probable

scour depth along the guide bund can be the worst abnormal
scour to be found in the river which is expected to be 2 to 2½
times the normal flood depth.

Gales proposed scour depth estimation to be based on

observed scour along eroding bend and increasing this depth
suitably for difference in severity of attack as on guide bund head,
body or tail and other factors. Gales recommendations in this
respect are reproduced in Table 5.9.

90
 Table 5.9 : Estimation of Depth of Scour according to
Gales

River Discharge (m3/sec)

Item 7086 to 21254 to 42507 to
21254 42507 70847

Observed deepest scour below low water x1 x2 x3

level along a soft cutting bank in the bend at
¾ falling flood.
Add 33 percent to convert these depths to 0.33x1 0.33x2 0.33x3
those obtainable at a rigid bank
Deepest known scour 1.33x1 1.33x2 1.33x3
Percentage addition to deepest known scour
to be made for contingencies such as unlike-
lihood of finding absolutely deepest scour,
narrowing of the river & for severe attack on
the guide bund head.
For body and tail of guide bank 25% 32% 45%
For upstream head of guide bank 50% 63% 60%
Deepest known scour to be adopted below
low water level
For body and tail of guide bank 1.66 x1 1.75 x2 1.93 x3
For upstream head of guide bank 2.00 x1 2.17 x2 2.53 x3

Khosla proposed scour factors depending on the part of

guide bund under consideration as in Table 5.10.

Table 5.10 : Estimation of Depth of Scour according to

committee headed by Khosla

Location Range of Mean depth to

Scour Depth be adopted

Nose of guide bank 2.0 to 2.5 D 2.5 D

Transition from Nose to 1.25 to 1.75 D 1.5 D
Straight Portion
Straight Reach of Guide 1.0 to 1.5 D 1.25 D
Bank

91
 D is Depth below H.F.L. and proposed to be estimated
using Lacey formula

D = 0.473 (Q/f)1/3

referred to before. If width of river channel is found to be less than

that obtained by Lacey formula

W = 4.836 QI/2

depth formula in terms of Q becomes inapplicable and the one

based on discharge intensity was proposed to be adopted viz.

D = 1.34 (q2/f)1/3

wherein q is the discharge intensity in m3/s per metre width.

Inglis advocated that for large, radius nose of guide bund,

scour should be estimated using the formula.

D = 2.75 DLacey

implying scour factor of 2.75

Sethi gave different scour factors for different parts of

the guide bund as in Table 5.11.

Table 5.11 : Estimation of Depth of Scour according to Sethi

Location Scour Depth

below h.f.l.

Upstream nose of curved head 2.75 D

Straight portion’ of shank and tail 1.75 D

Portion of Shank opposite pier and 33 m 2.0 D

on either side

Note: D is flow depth in metres according to Lacey formula as

in Table 5.10

92
 "Bridge Sub-Structure Code" of Railways(5.10) specified
that depth of scour be estimated using the Lacey formula.

D = 0.473 (Q/f)1/3

The scour factor recommended for upstream curved head

was 2.5 to 2.75 and for the rest of the guide bund portion 1.5.

Considering recommendation of various authors in

respect of scour factors, provisions in 'Bridge Sub-structure Code'
appear sufficiently conservative.

5.3.5.3 Apron quantity

In working out apron quantity, thickness of pitching is

required to be increased to allow for difference in quality of work
when pitching is laid in dry and pitching formed by launching
under water after scour in river bed. Spring and Gales proposed
increase in pitching on this account as well as on account of
other factors as given in Table 5.12 and 5.13.

Table 5.12 : Thickness of Launched Apron according to

Spring

S. Item For For

No. Shank Head

1. Thickness of pitching stone on As in 1.25 T

slope ‘T’ Table 5.5

2. Additional thickness for

possible non-uniform launching:
(a) in fairly steady rivers 25% 25%
(b) in rivers liable to sudden 50% 50%
deep scour

3. Allowance for fanning at curved - As in

heads Fig.5.9

93
 Table 5.13 : Thickness of Launched Apron according to Gales

For River Discharge For River Discharge For River Discharge

7086 to 21254 m3/s 21254 to 42507 m3/s 42507 to 70847 m3/s
SN Item Up-stream Body and Up-stream Body and Up-stream Body and
Head (cm) Tail (cm) Head (cm) Tail (cm) Head (cm) Tail (cm)
1 Thickness of pitching stone 105 105 105 105 105 105
2 Addition for absence of soiling 35 35 35 35 35 35
at 33%
3 Addition at 11 and 22% for high 11 11 23 23
discharge

94
4 Addition at 11 and 22% for high 11 11 23 23
silt content
5 Addition at 22% for head 23 23 23
6 Total thickness of pitching 163 140 185 163 209 186
stone
7 Apron thickness = 1.5 times 245 210 278 245 315 278
thickness of pitching stone
over 2 horizontal : 1 Vertical
launched face
8 Berm thickness Same as apron Thickness
 At curved heads, the apron launches on a conical surface.
Apron quantity is accordingly required to be increased suitably
as in Fig. 5.9.

Volume of stone = 2.81 p D X T [r1 + 2(F + R) + D]

This quantity is to be laid in an area

2 3(
π r2 -- r2
2 )
5.3.5.4 Apron as laid

Dimensions of apron according to Spring and Gales are

indicated in Fig. 5.10(A) and 5.10(B). Gales and Sethi favoured
uniform thickness of laid apron while Spring advocated tapered
design. Spring's design has been found deficient near the toe
and hence uniform thickness appears superior. Sethi has omitted
berm but recommended extension of soling for 3 to 6 m beyond.
Width of apron when laid equal to 1.5 times the depth of scour
below riverbed has been found to permit satisfactory launching.
Since apron is assumed to launch on a side slope of 2:1, the
thickness of apron when laid is given by 'quantity of apron' /
'width of apron laid.'

5.3.5.5 Construction programme

Suitable phasing of construction of various components

is essential for successful completion of the project and efficient
functioning of the bridge. Construction of only part of the guide
bunds or only one of the two guide bunds in one working season
is considered inadvisable. Both guide bunds, complete with their
curved heads should be completed simultaneously in one working
season.

5.3.6 Provisions as contained in IRBM para 810 with

regards to Guide Bunds

5.3.6.1 Necessity

Guide bunds are meant to confine and guide the river flow
through the structure without causing damage to it and its

95
approaches. They also prevent the out flanking of the structure.

5.3.6.2 Shape and design features

a) The guide bund can either be divergent upstream or

parallel. In the case of divergent guide bund, there is
possibility of formation of a shoal at the center. Parallel
guide bunds minimise obliquity and separation of flow along
the flanks. According to geometrical shape, the guide bunds may
be straight or elliptical. In the case of certain type of alluvial
rivers with sandy bed and meandering pattern, elliptical shape
appears preferable to minimise obliquity and separation of flow.
Various types of guide bunds are shown in Annexure 8/3.

b) Normally the upstream shank of the guide bund is between

1.0 to 1.5 times the length of the bridge, while the
downstream shank is between 0.25 to 0.4 times the length of
the bridge.

c) The tail bund on the downstream side is provided to afford an

easy exit to the water and to prevent formation of vertical
whirlpools or rollers which give rise to scour. These tail bunds
are also curved at their ends and should be properly
armoured.

d) The guide bund is provided with a mole head on its upstream

side. The mole head bears the brunt of the attack and
should be provided with adequate protection in the form of
slope pitching and properly designed launching apron. The shank
i.e. the portion behind the curved mole head of the guide bund
should also be similarly protected on the river side. The
slope in the rear of the guide bund need not necessarily be
provided with pitching and may be protected by planting
grass or shrubs as found suitable.

e) Radius of curved upstream mole head may be taken as 0.45L

(L is water way width determined from Lacey’s formula subject
to minimum of 150m and maximum of 600m). The radius of
downstream curved tail may be kept as 0.3 to 0.5 times the
radius of upstream curved head. The angle of sweep of
curved head may range from 1200 to 1450 according to river
curvature and that of the tail head may be kept as 450 to 600.

96
For smaller rivers, one single radius is good enough. Forimportant
rivers, multi radii may beselected generally after model
studies for smoother flows
.
f) Top width of the shank of the guide bund should be wide
enough to permit plying of trucks and keeping reserve
boulders for maintenance. From this consideration top width
may be taken between 6m to 9m, and side slopes may be
taken as 2:1.

g) Side slopes of guide bund needs protection on following counts:-

i) Wave action on the upstream side

ii) Water current along the slopes
iii ) Wind action
iv ) Rain cuts/Rain water

Most common method is to provide stone pitching. It is

necessary to provide 20 cm to 30cm thick graded filter below
the pitching. Stone used for pitching is generally man size
boulder of 35 to 55kg so that they cannot be easily displaced by
the current. For small works, one stone thick pitching (25 to
30cm) should suffice. Gaps in between could be filled up by
smaller pieces.

In case of guide bund, the pitching should continue right up to

the top of the formation for the river side, including the curved
head on both sides and tail head. For important rivers or in case
of large ponding etc, the pitching should be done on the rear side
of the guide bund also. For approach embankment, on the
upstream side, the pitching should continue up to the free
board level which should be determined not only on HFL but
also to take care of velocity head (V2/2g), wave action etc.
For the downstream side, pitching may be done up to the
water level based on hydraulic model study or general water
level observed.

A good drainage is key for protection of slopes from rain cuts,

particularly on high banks of over 6m height. For this,longitudinal
and cross drains should be provided.

Guide bunds and approach embankments particularly in

97
khadir of the river must be constructed in one go in one season.
In case this is not possible, at least, a wedge size equal to
angle of internal friction of the old construction should be
removed and the next construction should be done with
proper benching. For slope protection and apron, an overlap may
be provided.

h) No spurs projecting from the guide bunds should, in any case,

be provided.

j) For design and construction of guide bunds/launching

aprons reference may be made to IS: 10751-1994 (Planning
and Design of Guide Banks for Alluvial Rivers – Guidelines)
and IRC: 89-1997 (Guidelines for Design and Construction of
River Training and Control Works for Road Bridges)

5.3.6.3 Apron Protection for guide bunds

a) Apron is provided beyond the toe of the slope of the guide

bund, so that when bed is scoured, the scoured face will be
protected by launching of the apron stone or wire crate
containing stone.

b) Following are the important details for design of apron:

i) Thickness of apron

Thickness of apron is governed by thickness of pitching on the

slopes of the guide bund (T). In case of straight portion of
guide bund, the thickness of apron through its width is generally
kept as 1.5T. In case of curved portion of guide bund, the thickness
of apron is generally kept as 1.5T at the junction of apron with
pitching on the slope and the same is increased through its width
to 2.25T at the end of apron.

ii) Level at which the apron is to be laid Normally apron should be

laid on dry bed, as low as possible.

iii) Width of apron

Width of apron is determined by depth of scour and is generally

98
kept as 1.5 times the differen ce between the deepest known
scour level and low water level.

5.3.6.4 Maintenance:

a) Substantial reserve of pitching stone should be maintained on

the guide bund for use during emergency. This should be stacked
at the top of the guide bund. Quantity of reserve stock to be
maintained at guide bund should also be specified by Principal
Chief Engineer/Chief Bridge Engineer.

b) The track on the guide bund, where provided, should be

maintained in a satisfactory condition and should be capable of
taking boulder trains at any time. The Permanent Way Inspector
and the Assistant Engineer should inspect the track soon
after the monsoon every year and carry out necessary
repairs well before the next monsoon.

c) Every effort should be made to ascertain whether the

apron is launching to the intended position and this should be
done by probing after the flood season is over. Plotting of the
levels will indicate the efficacy of the launching.

d) Disturbance of pitching stone on the slope indicates dangerous

condition and additional stones should be placed in position
immediately as necessary.

5.3.6.5 Failures and remedial measures:

The conditions under which an apron of the guide bund can

fail and remedial measures to be adopted are stated below:

a) If the launching takes place beyond the capacity of the stone

in the apron and results in leaving the bank material exposed
to the current and wave action, more stone will have to be
added to the apron.

b) If stones are carried away by high velocity current from

the launching apron and the toe of the bund, the apron should
be strengthened against severe attack by laying large sized
stones at the outer edge of the apron.

99
c) If slips and blow -outs in the bund occur due to a steep sub
soil water gradient resulting from a rapidly falling flood in the
river, the bank should be widened to reduce the hydraulic
gradient. This equally applies to marginal bunds.

d) Wherever disturbance is noticed in rear of guide bund

due to wave lash or other causes, the slope pitching should be
adopted as a remedial measure.

e) An apron can launch satisfactorily only if the material scours

easily and evenly and the angle of repose of the underlying material
is not steeper than that of the stone. In all these cases action
should be taken to dump the boulders on the toe of the bank and
make up irregular surface.

5.4 SPURS OR GROYNES

5.4.1 Functions

Spurs or groynes are structures constructed in the river

transverse to the bank to achieve any of the following objectives.

Deflecting spurs are commonly used for protection of

riverbank from erosion either in a straight or a curved reach. When
navigation channel has inadequate depth, spurs are provided to
constrict the channel width so that depth is increased suitably
by bed scour. Closure of bye channels may be required to be
made in the interest of navigation or as a river training measure.
It is possible to achieve this purpose by means of spurs. In the
approaches to bridges and water intakes, shifting position of the
channel is some times required to be stabilised. Such stabilisation
can be effected up providing attracting spurs.

5.4.2 Types

Spurs can be of varied types. They may be either

permeable or impermeable. Permeable spurs are open structures
constructed by driving wooden balties filled in with brush wood
and weighed down by stone. When concentration of suspended
sediment load is heavy, permeable spurs cause quick siltation
due to damping of velocities. Such spurs are thus helpful in
protecting the bank by forming sediment berms along the toe.

100
 APRON

SLOPE
TOP

PLAN

F
HFL

R LWL
r1
AT BED LEVEL
r2 D

r3 SCOUREDBED

r4

SECTION

Fig. 5.9 : Apron for curved headof aguide bund with

allowance for fanning out

101
 1.5 D
H.F.L.
T
F a 2.76 T R L.F.L.
GUIDE BANK
SECTION
D
PROBABLE SECTION OF
WORST ATTACK

AVERAGE THICKNESS DEEPEST KNOWN SCOUR

1.25 T

ACCORDING TO SPRING

F = Free board.
R = Rise of flood.
D = Deepest known scour.
T = Thickness of slope stone.
Area of slope stone = 2.25 T (R+F) ie. T (R+F) Cosec a
Area of apron stone = 2.82 DT ie. TD Cosec a
Width of apron = 1.5 D.
Mean thickness of apron = 1.88 T.
Inside thickness of apron = T.
Outside thickness of apron = 2.76 T.
Inclination of slope stone = 2 to 1.
Desired inclination of apron stone = 2 to 1.

Fig. 5.10 (A) : Apron as laid according to spring

102
 DESIGNED THICKNESS
OF DEVELOPED APRON
PERMANENT SLOPE BERM APRON 1.5 D
E H.F.L.
P
LO R R1
C KS 1.5T1
BA T2 INCLINA- D3
T S C
GUIDE BANK TION
1 D1
SECTION T1 1 IN D D2
2
PROSPECTIVE SLOPE 1 IN 2
LENGTH OF SLOPE 2.24 D
DEEPEST KNOWN SCOUR + ALLOWANCE

ACCORDING TO GALES

F = Free board.
R = Rise of flood above bottom of apron.
D = Depth of scour for calculation of apron stone.
T = Thickness of permanent slope stone.
S = Thickness of soling.
T2 = Thickness of covering T + S.
C = Thickness of clay covering.
T1 = Thickness of stone on prospective slope
bottom of apron.
W = Width of berm = 15 for class A Rivers.
= 20 for class B Rivers and
= 25 for class C Rivers.
Area of permanent stone = 2.24 (R+F)
Area of prospective slope stone = 2.24 D x T1
Area of berm stone = W x 1.5 T1
Width of apron = 1.5 D.
Thickness of apron = 1.5 T1.
Area of apron stone = 1.5 D x 1.5 T1
In construction abrupt changes in the width of the
apron should be avoided.
Back slope to be suitably protected by stone
pitching or grass.
REFERENCES :
Class A Rivers - Q between 2.5 and 7.5 lakh Cs.
Class B Rivers - Q between 7.5 and 15 lakh Cs.
Class C Rivers - Q between 15 and 25 lakh Cs.

Fig. 5.10 A : Apron as laid according to gales

103
Impermeable spurs are made of solid core with exposed faces
protected by pitching. Such spurs can withstand severe river
attack better than permeable spurs. Deposition occurs on the
downstream side of the spur due to material scoured from the
nose and the high velocity current gets deflected away. Solid
spurs are accordingly provided where attack of flood flow is to be
diverted. According to height, the spurs are classified as full height
spurs when their top level is designed to remain above the highest
flood level. On the other band, when top level is lower than the
high flood level, spurs get occasionally submerged. Such spurs
are called part height spurs. Short submerged spurs with height
not exceeding 3 m or so when suitably aligned with respect to
flow direction are effective in protecting the bank and are termed
as bed bars. Another scheme of classification is according to
orientation of the spur. Spurs may be aligned with respect to the
flow direction, facing either upstream, normal or downstream.
Spurs facing upstream are termed deflecting or repelling whereas
spurs facing downstream are called attracting spurs. The term
attracting is used in a restrictive sense. Attracting action of the
spur can only be limited and that too under favourable conditions.
Spurs aligned normal to direction of flow are called normal spurs.
According to the shape of the spur head, the spur is termed a
T-headed spur, a hockey stick spur, a curved headed spur or a
round nosed spur. Material of construction is still another criterion
used in differentiation of spurs. Spurs may be either stone spurs,
brick spurs, bally pile spurs, spurs constructed of stone crate,
concrete blocks or trees.

5.4.3 Location

Spurs are required to be properly located to achieve the

desired purpose. When they are constructed for bank protection,
eroding reach needing such protection is required to be
determined by study of successive bank lines. In straight reaches
of rivers, bank erosion may be caused by excessive velocities
during high flood, obliquity of flow or squeezing of channel section
by moving islands in a braided river. In meandering rivers, erosion
is caused due to increase of depth, steepening of side slope and
consequent instability of bank along concave side of a bend,
which results in progressive erosion.

104
5.4.4 Length, spacing, inclination and height

Minimum length of a spur should be such that extent of

scour hole at the nose as well as embayment in between the spurs
do not reach the bank. Extent of scour hole in plan in the direction
towards the bank can be assumed to be 2.75 times the scour
depth D Lacey. Estimation of scour depth is dealt with in the
Chapter 6. The embayment in between the spurs depends on
spacing of spurs, bank alignment and orientation of spurs. In a
straight channel, for normal spurs, the embayment is roughly
equal to length of the spur when spacing between the spurs is
three times the spur length, as has been observed in Kosi river
downstream of Bhimanagar barrage. In an acute bend, embayment
equal to length of the spur is obtained with spacing of spurs
roughly equal to twice the spur length, as observed in Ganga river
in acute bend at Mansi. If spurs are too long, the constriction of
river section may become excessive and the attack on spurs can
be severe. In such a case the velocities and channel depth on
the opposite bank may also get affected. In any case extending
the spur length to obstruct deepest course carrying main flow of
the channel is generally hazardous. Channel obstruction beyond
20 per cent of the width is normally avoided. Length and spacing
are related to economics, the nose portion costing heavily. Fewer
number of spurs with required longer lengths of shanks, consistent
with considerations already mentioned, therefore, provide an
economical arrangement. Possibility of bank erosion occurring
in between the spurs at river stages lower than the design flood
stage needs also to be kept in view. For bank protection, spacing
equal to 2 to 5 times the projected length of the spur is normally
used.

Angle of the spur with direction of approach flow governs

scour and formation of separation eddies. When the spur faces
upstream, the scour is severe at the nose but scour along shank
is much less. Separation eddies are formed both on the upstream
and the downstream. Material scoured at the nose forms a shoal
by deposition in the eddy portion on the downstream. Such spurs
are called deflecting spurs.

With spur facing downstream, the flow hugs the

upstream face of the spur without separation and the scour at
the nose is less. The downstream eddy and shoal however still

105
form. The length of bank protected on upstream is shorter but on
the downstream, protected length measured from the point of
the junction of the spur with the bank is increased. When a battery
of spurs is provided, the first spur has to take brunt of the attack
and this spur is often oriented facing downstream. Rest of the
spurs are made either normal or facing a little upstream. The
angle of spurs to the flow direction in case of upstream or
downstream facing spurs is ordinarily limited to 20 degrees with
respect to normal, upstream inclination being favoured especially
in curved reaches.

Spurs are normally made full height with top level above
design high flood level providing· an adequate freeboard. When
high ground to tie the spur is not available within a reasonable
distance, the top level of the spur can be same as the bank level,
special care being taken in this case to prevent outflanking by
breach near the bank. Downstream face of the spur and the
riverbank in the vicinity of the spur need adequate protection.
Partial height spurs have been used with success in some
countries. Normally, however, full height spurs are preferred.

Length, inclination and spacing of spurs at important

locations are finalised preferably by resorting to hydraulic
modelling.

5.4.5 Materials of construction

Shank of a spur is constructed of sand core and its

exposed faces are protected by stone pitching. The attack at
nose is very heavy and hence the nose may be constructed
wholly in stone. Alternatively if attack is heavy all over the spur
length or if it is easier and not very expensive to construct the
spur wholly in stone, such a construction may be preferred. The
apron should be in stone of the same size as used for slope
protection of guide bunds.

Permeable spurs are constructed by driving wooden bally

piles to achieve penetration below riverbed of 6.0 to 9.2m. The
ballies are normally 22.5 to 37.5 cm in diameter. Mattress or
stone carpet is provided on river bed to prevent scour.

106
5.4.6 Provisions as contained in IRBM para 811 with
regards to Spurs/Groynes

5.4.6.1 A spur/groyne is a structure constructed transverse

to the river flow and is projected form the bank into the river.

5.4.6.2 Type of Spurs/Groynes

i) They may be either “Permeable” or “Impermeable”.

Permeable spurs are constructed by driving wooden bullies or
bamboos, filled in with brush wood, with sarkanda mattresses
or other suitable material. These are helpful in causing quick
siltation due to damping of velocity. They are useful when flood
velocities 10 are relatively lower and concentration of suspended
sediment load is heavy.They allow water to pass through.
Permeable structures are discussed in detail in Para 811(5).
Impermeable spurs are made of solid core, constructed of stones
or earth and stones with exposed faces protected by
pitching. These spurs can withstand severe attack better than
permeable spurs.

ii) Spurs may be classified as (a) repelling (deflecting)

(b) attracting and (c) normal (sedimenting). Repelling (deflecting)
spurs are those which incline upstream at an angle of 60
degree to 70 degree to the river course and deflect the current
towards the opposite bank. They cause silting in still water
on the upstream pocket. Attra cting spurs incline downstream
and make the deep channel flow continuously along their
noses. They cause scour just on the downstream side of
the head due to turbulence. The river flow is attracted
towards the spur. Normal (sedimenting) spurs are those whi
ch are built at right angles to the bank to keep the stream in a
particular position and promote silting between the spurs. They
have practically no effect on the diversion of the current and
are mostly used for training of rivers for navigational purposes.

iii) Spurs are also classified as full height spurs and part
height spurs. Where top level is higher than HFL, it is called a
full height spur.

iv) Spurs are also constructed extending into the

stream with a “T” head or hockey stick shaped head, properly

107
armoured to hold the river at a distance. A series of such spurs/
groynes correctly positioned can hold the river at a position away
from the point intended to be protected. The edge of the "T”
head should be curved somewhat in the manner of a guide bund
to avoid swirls.

5.4.6.3 Location and salient features of a Spurs/Groyne

i) The space between spurs or groynes generally

bears a definite ratio to their length. The common practice is
to keep the spacing at about 2 to 2.5 times the length so as to
effectively protect the bank.

ii) If designed as a full height spur, care should be

taken to see that spurs are built sufficiently high so that
they are not overtopped and out flanked by the current
during high floods. Free board of 1 meter is provided.

iii) The side slopes of spurs are generally 2:1.

iv) The spurs should be anchored on to high ground.

v) The head of the spur is most vulnerable point for

scour and should be well protected on slopes by pitching and
at toe by an apron designed for scour depth of 2.5 to 2.75
times DLacey at the mole head. For computation of DLacey,
Clause 4.6 of ‘IRS Code of Practice for the Design of Substructures
and Foundations of Bridges’ may be referred.

vi) Spurs should never be constructed at a point

where severe attack is taking place but at some distance
upstream.

vii) Spurs/groynes should be used only in situation

where they are absolutely necessary.

viii) The design of spurs may be finalised preferably

through hydraulic model studies.

ix) For design and construction of groynes (spurs)/

launching aprons reference may be made to IS:8408-1994
(Planning and Design of Groynes in Alluvial Rivers –

108
Guidelines) and IRC:89-1997 (Guidelines for Design and
Construction of River Training and Control Works for Road
Bridges).

5.4.6.4 Maintenance of Spurs/ Groynes

In all cases, satisfactory arrangement should be

made for the maintenance of spurs/groynes by providing access
to them during all seasons of the year and keeping boulders
as reserve. The maintenance procedures specified for guide
bunds apply equally to spurs/groynes also.

5.4.6.5 Permeable structures

a) Permeable structures can be used either

independently or with the support of other impermeable stone
structures or river training and bank protection measures.
These structures are easy to construct, use low cost locally
available material and require limited skill in construction.
These are very handy in anti-erosion works during emergencies
in floods. These structures can also be used in areas where
good quality stones are costly and/or not available. Thus
permeable structures are cost effective alternative to the
river training or anti-erosion works with impermeable spurs.
Depending upon the purpose to serve, the permeable
structures are constructed transverse or parallel to the direction
of flow. Permeable structures serve one or more of the following
functions:

i) Training the river along a desired course.

ii) Reducing the intensity of flow at the point of river attack.
iii) Creating a slack flow to induce siltation in the
vicinity of the permeable structures and in the
downstream reach.
iv) Providing protection to the bank by dampening the
velocity of flow along the bank.
b) The permeable structures can be classified as follows:
i) According to function served, namely, diverting and
dampening, sedimenting.
ii) According to the method and material of
construction, namely, bally, bamboo, tree and willow

109
 structures.
iii) According to the conditions encountered, namely,
submerged and non-submerged.
iv) According to the type of structure provided, namely,
spur type, screen type or dampeners (revetment)
type.

c) The permeable structures are made up of different

types of smaller units called elements. Many elements, made
up of bamboos, ballies, RCC poles etc. are arranged in specific
pattern and linked together to form a permeable structure.
Different types of elements used for making permeable structures
are as following:
i) Porcupines–Porcupines are typically made up of
bamboos/ballies, have cubical/prism shaped box at
the central portion with their legs extending in all
directions. The overall size is 2m to 3m. The central
box is filled with stones for stability of individual unit
during floods.
ii) Cribs– This is a pyramid type of structure made
up of bamboos/ballies with a box at the bottom for
holding stones for stability during floods. Size of the
box is generally square of size 2m to 2.5m at the
bottom. Total height of the structure is 3m to 4m.
iii) Bally frames -Permeable bally structures are made up
of main skeleton of large bamboos or ballies. Cross
ballies are used for stability of the structure.
iv) Tree branches– Branches of trees or trees of
short height are hanged from a wire rope duly weighted
with stones and are aligned as a spur projecting into
the river. The wire rope is duly anchored on the bank
and in the riverbed.

d) The main criteria for the selection of the material

are cost and easy/local availability. Standard, commercially
available bamboos of girth 20cm to 30cm are used for the
porcupines and cribs. Smaller girth of 20cm to 25cm is used for
bracings. Standard, commercially available ballies of girth

110
15cm to 25cm are used for the bally structures. Normally, the
larger girth of 20cm to 25cm is used for the main members,
whereas, the smaller girth of 15cm to 20cm is used for
bracings. Generally, 4 to 5 strands of 4mm GI wire are used for
interconnecting porcupines, cribs, and anchor them to the ground.
Ballies driven into the ground upto a depth of 2m are
treated as anchor. Concrete anchors have an anchor rod of
size 32-36mm, well embedded in concrete cube. Wire crate
anchors are of size 1.5m x 1.5m x1.5m, made up of thick
wires and filled with stones or bricks. A concrete block is
casted with bolt and is included in the wire crate anchor. In
case of emergencies, tie wires are joined directly to the wires
of the crates.

e) In case of shallow water flows and upto maximum

depth of flow 3m to 4m, porcupines are used for both spurs
and screens. For maximum depths of flow from 4m to 6m,
cribs are preferred. For the depths beyond these limits,
bally spurs are preferred.

f) Permeable structures commonly used are spurs,

dampeners and screens.
i) Spurs are generally made up of 3 to 4 rows of
porcupines or 4 to 6 rows of cribs. Schematic sketch
of typical permeable spur is shown in Annexure-8/
6(d). On a straight reach, permeable spurs are
normally spaced at 3 to 4 times its length. On a
curved channel, depending upon the obliquity of flow,
the spurs are normally spaced at 2 to 3 times the
length. Projection of the spurs into the river channel is
normally 11% to 15% of width of channel. Three
spurs are normally provided for a specific reach to
be protected. A single permeable spur is generally
not found effective. Alignment of spurs is kept
pointing towards upstream.
ii) For depth of flow up to 3m, two rows of porcupines
are laid along the banks on either side at the toe as
dampeners. For more depth, numbers of rows are
increased.
iii) Permeable screens are used for choking the secondary
channels. 4 to 6 rows of porcupines or 6 to 9 rows

111
 of cribs are normally used in a permeable screen.
One screen is normally provided at the entrance
of the bypass or secondary channel. The second
screen is provided at a distance of 1 to 1.5 times
width of the screen and is extended on both the
banks for a length one third of the channel width.

g) Due to inherent weakness of the elements, the counter

weights are provided in the central box of the porcupines or in
the bottom tray of the cribs. Due care is necessary to tie the
weights to the main body of the elements. The elements are tied
to each other by wire ropes. The tie ropes are duly
anchored to the bank and at the nose with the help of suitable
anchor or anchor blocks. Intermediate anchors are also provided
at an interval of 15m to 20m along the length of the structures
on the upstream side.

h) No bed protection is needed for the structures

made up of porcupines and cribs. Sinking of these structures
into riverbed is a welcome feature, which adds up to the stability
during floods resulting in better performance.

5.5 APPROACH BANKS AND MARGINAL EMBANKMENTS

5.5.1 Approach banks

Since country slope is towards the river approach banks

are usually required to be provided on either side of the bridge.
When the approach cuts off, whole or part of spill discharges,
parallel flow is developed along its upstream face, and at the
abutments deep scour holes are formed due to concentration of
obstructed discharge. In alluvial rivers when bridges are provided
with constricted waterway, guide bunds are normally added. The
river can then form a single or double loop behind the guide bund.
If the guide bund is not sufficiently long, such an embayment
can cut into the approach bank. Even if length is adequate to
keep the embayment away from the approach bank, parallel flow
may still be obtained during floods. Adequate protection is required
to be given to the upstream face of approach bank against all
such eventualities.
The approach banks need to be made sufficiently high
since during floods the water level on upstream side is governed

112
by the high flood level, which is further raised by the approach
velocity. On the downstream side, the water level may be very
low. Under these conditions, the approach bank has to withstand
high differential head. Rapid draw-down during falling flood can
create further, instability in the approach bank. Nature of
foundation, material used in forming the approach bank, cross
section and protection on side and at toe are important aspects
governing stability of an approach bank. Cross section of the
bank and surface protection are accordingly designed considering
foundation strata, engineering properties of the soil forming the
bank and the hydraulic conditions which the bank has to
withstand. Necessary precautions are required to be taken, in
consonance with prevailing conditions. For instance, in order to
discourage formation of rat holes, inverted filter and suitable other
means may have to be resorted to.

Alignment of approach has relevance in the context of

possible embayment forming at the back of a guide bund and
likelihood of active channel developing along the approach bank.
Approach bank alignment deviating downstream towards the khadir
edge is helpful to reduce the attack under these conditions. When
approach banks are considered vulnerable or threatened by river
attack, an advance low bank termed as sub bank is sometimes
provided to act as the first line of defence.

When river khadir is wide, approaches become longer

and water can stagnate behind the guide bunds over long lengths.
It is inadvisable to provide culverts or secondary bridge openings
with liberal capacity to drain this water. Such openings may
encourage the river to develop a direct course through them which
is dangerous.

5.5.1.1 Provisions as contained in IRBM para 817 with

regards to protection of Approach Banks

1. Approach banks of bridges may be subjected to severe

attack under the following conditions:
i) When the HFL at the bridge is very high and there is
spill beyond the normal flow channel.
ii) When the stream meets a main river just downstream
of the bridge.

113
 iii) In the case of bridges with insufficient water way.
iv) The wave action on the approach bank of bridges
situated in a lake/large tank bed may have a detrimental
effect.
In all the above cases the pitching of the approach bank up to
HFL with sufficient free board is an effective solution. Provision
of toe wall and narrow apron in some cases will also be useful.

2. If deep borrow pits are dug near the toe of approach banks,
the water flows through these pits and forms a gradually
deepening water course which may eventually threaten the
safety of the approach bank. In this case it will be useful to put
rubble “T” spurs across the flow to reduce the velocity and
expedite silting of the course.

3. Whenever the water level on either side of an approach

bank is different, there may be seepage of water and to ease
the hydraulic gradient, widening of banks, provision of sub
banks and toe filters etc may be resorted to.

4. At locations with standing water against the embankment,

special watch should be kept when the water level recedes rapidly
and when slips are likely to occur.

5.5.2 Marginal Embankments

When railway lines are located on both the banks of a

river as in case of the Ganga, the railway banks act as marginal
or flood embankments. Since the spill depth can be considerable,
height of these banks is usually large. When river is in flood,
spills may develop parallel flow along riverside face of the marginal
banks. Differential head across these banks can act in both
directions depending on the lag in timings and duration 0f floods
in the parent river and tributaries.

If marginal embankments are located too near the khadir

edge, changes in river meanders may endanger their safety by
causing breaches. Alignment and section of marginal
embankments have to be properly designed keeping in view all
the factors.

114
 Effect of marginal or flood embankments is to restrict
the spill discharge and divert it to the main channel. Discharge
intensity in the river channel is accordingly increased which
improves its sediment transporting capacity. The riverbed is,
therefore, normally lowered. Such a lowering is however not
reflected in flood levels since confinement for spill flow results in
boosting up of water level as is exemplified in case of the
Mississippi river. If the river is initially of aggrading type
construction of flood embankments may not be effective in
neutralising this tendency and the river may continue to aggrade
though at a slower rate.

5.5.3 Provisions as contained in IRBM para 812 with

regards to Marginal Bunds

Marginal bunds are provided to contain the spread of the river

when the river in flood spills over its banks upstream of the bridge
site over wide area and likely to spill in the neighbouring water
courses or cause other damages. The marginal bund should
normally be built well away from the active area of the river.
The slope should be well protected by turfing. Where a marginal
bund has to be built in the active area of the river, it should be
protected with pitching and apron. The earth for the
construction of marginal bund should preferably be obtained
from the river side. The upper end of the marginal bund
should be anchored into high ground well above HFL. Marginal
bunds should be inspected every year along with the annual bridge
inspection and necessary repairs should be carried out before
the onset of monsoon. Cattle crossing and rodent holes
across the marginal bund should be specially watched and
deficiencies made good.

5.6 BANK REVETMENT

Eroding bank of a river can be protected by constructing

spurs or by means of bank revetment. Spurs cause constriction
of river width, heavy scour and turbulence at noses, eddies on
upstream and downstream side and may involve prohibitive cost.
In comparison, bank revetment may be found to be economical
and does not normally affect river regime. Bank revetment is,
therefore, sometimes found preferable to spurs. Disadvantage of
a revetment is that it creates a deep channel along its toe and

115
thus tends to pin down the channel position which may not be
desirable under certain conditions. Secondly the eroding current
brushes past the pavement and is not diverted away from the
bank as in case of spurs.

Revetment can be of various types. Most common

revetment is with stones, either handset or machine placed, riprap
or stone in crates. Brick pitching may be laid on edge or pitching
may be formed of brickbats in crates or of brick blocks.
Alternatively pitching may be formed of soil cement blocks,
concrete blocks. concrete slabs or artificial body forms like
tetrapods. Articulated concrete mattress is extensively used on
the Mississippi river. Pavement formed of interconnected concrete
slabs is also popular and commonly used in European countries
and the U.S.S.R. Asphalt mat, asphalt carpeting, lumber
mattress, bamboo mattress, fascine mattress also provide good
surface protection though their life cannot be as long as of stone
or concrete pavement. In emergency works gunny bags filled
with sand or ballast from track are used when stone is not
available. Plastic bags filled with sand and properly sealed are
preferred as their life is longer. However several patented systems
of erosion control mattresses of synthetic fabric are available
and are far too superior. Different types of prefabricated synthetic
fabric forms are in vogue which can be laid at site and then filled
by pumping mortar. Polyethylene tubes upto 2m diameter size,
hydraulically filled with sand after placement at site have also
been used successfully (5-15).

Revetment is laid on riverbank after grading the side

slope. Eroding banks often acquire quite a steep side slope
whereas the natural angle of repose of pitching material may be
much smaller. It is necessary that the side slope of the bank be
flatter than the angle of repose of material used for pavement. In
plan kinks and protrusions in the bank alignment are required to
be removed before laying pitching.

Size and thickness of pitching stone is governed by

velocity incident on it and the side slope on which it is laid.
Allowance for under water work is required to be made suitably.
Adequate toe protection is necessary to ensure safety of slope
revetment against undermining and slipping due to toe erosion.
In this context, possible bed scour is required to be properly

116
assessed. Stability of pitching depends on provision of well-
designed filter below it. All these aspects have been previously
dealt with under slope and apron protection for guide bunds.

5.7 CLOSURE BUNDS

Spill channels or bye channels of a river are often required

to be closed to obviate possibility of attack on bridge approaches
or marginal banks. Such closures can be effectively and
economically achieved by means of permeable screens of
construction similar to permeable spurs. When such closure work
is across the channel carrying substantial discharge, penetration
of piles, apron protection, bracings along and across flow
direction, stay supports, etc. are required to be designed with
adequate factor of safety.

5.7.1 Provisions as contained in IRBM para 813 with

regards to Closure Bunds.

Sometimes it may be necessary to entirely block one

or more channels of the river in order to prevent the discharge
of such channels developing into a main river channel after the
construction of the bridge. This is done by providing a closure
bund. The bund is designed as an earthen dam. The same is
generally constructed at some distance from the Railway
line. Special care should be exercised to guard it against its
failure. It should be inspected every year after the monsoon
and necessary repairs carried out.

5.8 ARTIFICIAL CUT OFFS

Natural cut offs occur in meandering rivers across

bottlenecks, especially in hairpin bends, due to progressive bank
erosion. In Punjab rivers ratio of bend to chord length when cut
off occurred has been roughly found to be 1.7 and in Indus about
2.0.

Cut offs are some times required to be brought about

artificially. Sharp bends in rivers hamper navigation and cause
shallow bars and oblique crossings to develop. Navigation is
immensely improved if such bends are eliminated by making cut
offs as successfully done in Mississippi river.

117
 Artificial cut offs across hairpin bends are quite useful
as a flood control measure. Consequent upon making a cut off,
the river slope on the upstream steepens, velocities increase,
bed scours and flood levels are lowered. In the Mississippi lowering
of flood levels by 1 to 2 m is achieved by means of cut offs.
Increase in velocity, however, results in accelerated bank erosion
and consequent change in shape and position of other existing
meanders on the upstream and bank protective measures are,
therefore, simultaneously warranted.

Cut offs are useful also as a river training measure.

Barrages are often constructed in dry and river subsequently
diverted to flow through the structure. Such diversions have been
found to be aided to a large extent by providing artificial cut offs.
When bridges are constructed on braided rivers, discharge
distribution in the approach to the bridge may· be very non-
uniform. Conditions at the bridge can be markedly improved by
effecting proper distribution of the flow by means of suitable cut
off across islands. Approaches to pump intakes also at times
necessitate making cutoffs to establish direct connection with
an active channel to ensure adequate water supplies.

5.8.1 Provisions as contained in IRBM para 814 with

regards to Assisted / Artificial Cut- Offs

Sometimes when very heavy meandering develops near bridges

and there is a danger of its encroaching too heavily into the
still water area or otherwise dangerously approaching the
Railway embankment, it becomes necessary to dig a cut-off
channel which will ultimately develop and help in the diversion of
water through it. To effect economy, a pilot channel cut is
usually made when there is low flow in the river and full
development of the channel takes place during the flood. This
cut-off channel should preferably have (i) at
least three times the river’s straight regime slope and (ii) the
upstream end should take off from where the bed load of main
channel has less than the average amount of coarse material i.e.
from the active part of the channel where the velocity is more. The
entrance to the pilot cut should be bell shaped to facilitate
entry of water. The chord loop ratio should normally be greater
than 1 to 5 if a successful channel is to develop. Cut off should
be planned with care taking all relevant factors into account.

118
5.9 PROTECTION FOR SHALLOW PIERS

Bridge piers offer obstruction to flow. Velocity field round

them is accordingly affected. On upstream side of the pier,
stagnation head is built up, the kinetic head changing to potential
head. Water level therefore, rises at this point, along sides, the
flow concentrates and velocities increase and water level drops.
The resulting pressure and velocity distribution develops a
horseshoe shaped vortex which causes maximum scour at the
upstream end of the pier. The scour at sides is somewhat less
and at the tail end it is least. Depth of scour is governed by
scouring action as well as sediment feed into the scour hole.

Scour depth around piers changes with river discharge.

During floods, scour depth may be large whereas during low water
season it may be small. In bridge design, flood scour is important
since piers have to be designed to be sufficiently deep below
bed level obtained after flood scour.

Devices have been tried to arrest the pier scour such as

rings and projections round the piers to break the pressure
gradient responsible for generating scour hole. These methods
are, however, found to be not sufficiently effective in the case of
major bridges. Adoption of piles or pile clumps for reduction of
scour is also not in popular use. The normal approach in designing
piers is to sink wells deep enough to ensure safety against worst
possible scour. In case of shallow piers of existing bridges, suitable
protection is given to ensure safety, as given in Chapter 6.

5.10 TRAINING AND PROECTION WORKS IN HILLY AND

SUBMONTANE RIVERS

In hilly reaches of rivers, the current velocities are fast

and the training and protection works are, therefore, required to
be massive and strong. Either stone crates or concrete blocks
of sufficient size and weight are used in such works. Interlacing
of crates and blocks may become necessary under exceptional
circumstances. Current with high velocity can negotiate only a
flat curvature and longer length is required for deceleration. Training
works are accordingly required to be more extensive and
streamlined.

119
 Landslips are common and hence precautionary
measures for efficient drainage and stability of bank slopes are
warranted.

For foundation, rocky strata may often become available

and hence, foundation difficulties may not pose a problem.
Similarly in hilly reaches scour problem may not generally be
required to be faced.

In submontane region, the channel pattern in normally

braided and training works are required to close several minor
and bye channels. Major channel may be prone to shift laterally.
Heavy and numerous training and protection works are, therefore,
required in this reach. In view of high velocity, design
considerations are similar to those in hilly reaches. In addition,
aggradation may be associated with tendency for building of an
alluvial fan, in which case rate of aggradation is required to be
examined. If this rate is fast, possibility of controlling it by means
of constriction of channel section or by reduction of sediment
load is required to be assessed.

REFERENCES
5.1 Spring, F.J.E. 'River training and control on guide bank
system', Technical Paper No. 153, Government. of India,
1903.
5.2 Gales, R., 'The principles of river training for the railway
bridges and their application to the case of Hardinge
bridge over the lower Ganges at Sara', Journal of the
Institution of Civil Engineers, December 1938, Paper
No. 5167.
5.3 Lacey, G., Stable channels in alluvium', Paper No. 4736,
minutes of proceedings of the Institution of Civil
Engineers, Volume.229, January 1930.
5.4 Sethi, H. K. L, ''River training and control for bridges"
Technical Paper, Research Designs and Standards
Organisation, Lucknow - 1960.
5.5 'Report of the Committee of Engineers, Government of
India, Ministry of Railways, October 1959.
5.6 Sharma, H. D., Goel, P. K., Singh, V. K., 'Elliptical guide
bunds', Journal of the Institution of Engineers (India)

120
 March, 1976. Vol. 56, part C 15.
5.7 Garg, S. P., Asthana, B. N., Jain, S. K., 'River training
at bridges and barrages,' Journal of the Institution of
Engineers (India), Volume 51, May 1971.
5.8 Chitale, S. V., 'Radius of curved heads of banks', Irrigation
and power, Journal Central Board of Irrigation Power,
India, Vol. 37, No.4, October 1980.
5.9 Inglis, C. C., 'The behaviour and control of rivers and
canals', Research publication No. 13, Central water,
Irrigation and Navigation Research Station, Poona, India,
1949.
5.10 'Indian Railway Standard Code of practice for the design
of Sub-structures of bridges (Bridge Sub-structure Code)'
Ministry of Railway, Railway Board, Government of India,
New Delhi, Revised 1985.
5.11 "Use of rip rap for bank protection", Hydraulic
Engineering Circular No. 11, June 1967, U.S. Dept. of
Transportation, Bureau of Public Roads, Federal
Highway Administration.
5.12 "Hydraulic Design of flood control channel", Engineer
Manual, 1st July, 1970, Department of the Army, U. S.
Army Corps of Engineers.
5.13 "Hydraulic Designs of Spillways", Engineer Manual, 31st
March, 1965, Department of the Army, U. S. Army Corps
of Engineers.
5.14 "Hydraulic Design Criteria", Sheets 712-1, U.S. Water-
ways Experiment Station Vicksberg, Mississippi, U.S.A.
5.15 Koerner, R. M. and Welsh, J. P., "Construction and
geotechnical engineering using synthetic fabrics", A
volume in the Wiley Series of Practical Construction
Guides, Editor Morris M. D., Wiley Inter Science
Publication, New York, U.S.A., 1980.

121
 Chapter 6

DESIGN OF TRAINING AND

PROTECTION WORKS

6.1 DESIGN DISCHARGE

6.1.1 Estimation of Design Discharge

Design discharge (Q) is the estimated discharge for the

design of the bridge and its appurtenances (accessories). This
shall normally be the computed flood with a probable recurrence
interval of 50 years (Q 50).

6.1.1.1 New Lines & Rebuilding of Bridges

Following methods are available for estimation of design

discharge (Q 50) as per Sub structure code Para 4.3

1. From actual data (SSC - Para 4.3.1)

Where stream flow records (yearly peak discharges) are available
for the desired recurrence interval or more, design discharge shall
be: The computed flood for the desired recurrence interval

2. Statistical methods (SSC - Para 4.3.2)

Where such Stream flow records exist for less than the desired
recurrence interval but sufficient for the statistical analysis, design
discharge may be: Computed statistically for the desired
recurrence interval

3. Unit hydrograph (SSC - Para 4.3.3)

Where records of floods are not available for sufficient length to

permit reliable statistical analysis but where Rainfall pattern &
intensity records are available for sufficient length of time & Where
it is feasible to carry out at least limited observations of rainfall &
discharge, Unit hydrograph based on such observations may be
developed and design discharge of the desired recurrence interval
computed by applying appropriate design storm.
122
4. Synthetic hydrograph concept & RDSO report RBF-16
(SSC - Para 4.3.4)
Where such observations, as mentioned in Cl. 4.3.3 above, are
not possible,
A synthetic unit hydrograph may be developed for medium size
catchment ( i.e. Area 25 sq. Km or more but less than 2500 sq.
Km) by utilising established relationships as mentioned in Flood
Estimation Report for respective hydro-meteorological sub zone.
For small size catchment (less than 25 sq. Km), design discharge
may be estimated using the techniques described in RDSO report
no. RBF-16, titled as “Flood Estimation Methods for Catchments
less than 25 km2 area.”

5. Other methods (stage-discharge relationship

(SSC - Para 4.3.5)
Where feasible, gauging of the stream may be done to establish
the stage– discharge relationships, Discharge at known HFL
determined. Otherwise, the discharge may be estimated by slope
area method after obtaining flood slope by field observations.
For Indian Rly bridges, the design discharge Q50 is generally
estimated on the basis of sub structure code para 4.3.4
above i.e. for catchment up to 25 sqkm using RBF-16 and
for catchment area 25 sqkm & above and up to 2500 sqkm
using flood estimation report (Synthetic Unit Hydrograph
concept)

6.1.1.2 Doubling Works & Gauge conversion Projects

G.C. & Doubling Projects (SSC - Para 4.5.7)

i) Where there is no history of past incidents of over flow/washout/

excessive scour etc. during last 50 years: The water way of
existing bridge may be retained after taking measures for safety
as considered necessary by Chief Engineer In charge.

ii) For locations where there is history of past incidents of over

flow/washout/excessive scour: The waterway has to be re-
assessed based on the freshly estimated design discharge using
clause 4.3.1 to 4.3.4 of sub structure code.

123
iii) For locations, where existing bridges are less than 50 years
old and there is no past history of incidents of over flow/washout/
excessive scour etc.: The water way may be judiciously decided
after calculation of design discharge and keeping in view the
waterway of existing bridges on adjacent locations on the same
river (Para 4.5.7of SSC)

Rebuilding of Bridges (SSC - Para 4.5.8)

For rebuilding of bridge: The waterway shall be determined

keeping in view the design discharge as worked out from clause
4.3 of sub structure code.

Gauge conversion projects (SSC - Para 4.5.8)

For strengthening existing bridges by jacketing etc., a reduction

in waterway area as per the limits specified below may be allowed
by the Chief Bridge Engineer provided that there has been no
history of past incidents of overflow/washout/excessive scour etc.
and that measures for safety as considered necessary by the
Field Engineer and approved by CBE are taken.
Reduction in Waterway Area
Span of Bridge Allowed as % Age of Existing
Waterway.
Up to and including 3.05 m 20%
3.05m to 9.12m including Varying linearly from 20% to 10%
Greater than 9.12m 10%

Further reduction in the area shall be subject to CRS sanction

and submission of detailed calculation of waterways etc. Where
the clearances are not available, the bridge should be rebuilt.

6.1.2 Design discharge for foundation (Q )

f
To provide for an adequate margin of safety, the foundation
and protective works of a bridge should be designed for a flood
discharge of higher magnitude than the waterway design
discharge (Q50). For this purpose, the waterway design discharge
(Q50) may be increased by the percentages given in Table. 6.1.

124
 Table 6.1 : Foundation Design Discharge (Q f )with
Percentage Increase to Waterway Design Discharge (Q50)

For rivers having catchment Percentage increase in waterway

areas at bridge site design discharge (Q50) to obtain
(Sq. km.) foundation design discharge (Q )
f

Upto 500 30%

More than 500 and 30 to 20% (Decreasing with

upto 5000 increase in area)

More than 5000 and 20 to 10% (Decreasing with

upto 25000 increase in area)

More than 25000 less than 10% (using discretion)

6.2 DESIGN OF WATERWAY

When the river flows between high banks and the whole
width is actively functioning during high floods, the bridge waterway
should practically be equal to the spread between stable banks
at the design flood level. This practice may be followed even if
such a river slightly overtops the banks during extraordinary
floods. These conditions may normally be met with in incised
rivers and in upland reaches in river gorges.

If in the above type of rivers, depth of spill is appreciable,

waterway should be suitably increased beyond bank to bank
width in order to carry spill discharge without causing afflux
beyond permissible limit.

In case of alluvial rivers, the bed is scourable. If width

at design flood stage in such rivers is more than the width
given by Lacey formula

P = W = 4.836 Q1/2

It would be desirable to constrict this width at bridge

section to achieve economy as well as improve the hydraulic
performance. According to Railway Bridge Sub Structure Code,

125
effective waterway for bridges on alluvial rivers should normally
be equal to width given by the Lacey formula.

P = 1.811 CQ1/2

wherein C is a coefficient normally having a value of 2.67 but

which may be varied from 2.5 to 3.5 to suit nature of bed material
and characteristics of flow. Higher value of C is applicable to fast
running streams.

Obstruction due to piers is allowed for by adding double

the sum. of the weighted mean submerged width of all the piers
including footings for wells to arrive at the total width of water
way to be provided between ends of the bridge according to Railway
practice. (6.7) Similarly one extra span is added to allow for
obstruction due to training works in the end spans.

If bank to bank width of an alluvial river is shorter than

Lacey width, the bridge should span the entire width but not be
extended further to make the waterway equal to Lacey width.

Even if the river is alluvial, but the flood is flashy, time

may not be sufficient to cause bed scour, constriction on basis
of Lacey width is then not permissible. In such cases maximum
depth of scour should be ascertained from field observations and
waterway determined by Velocity- Area method. If river data of
velocities at flood stages is not available, a suitable velocity formula
may be employed for this purpose. Progressive scour-afflux
computations may also be made for the design hydrograph as
explained in Chapter 4 and results used in finalising the design
waterway.

For rivers in submontane region, the slopes are steep,

velocities high, bed material ranging from heavy boulders to gravel
and floods often of flashy nature. The constriction in such rivers
should be governed largely by the configuration of active channel
or channels, the cost of diversion of channels and the cost of
guide bunds which can be much longer needing heavier protection
than in case of guide bunds in sandy bedded rivers.

In tidal reaches of rivers, the regime is predominantly

governed by tidal characteristics. Constriction is often likely to

126
affect the prevailing tidal pattern and cause deterioration by silting.
The possibility of constriction of a tidal waterway should, therefore,
be viewed from all angles including tidal regime, navigation
requirements, effect of salinity, etc.

6.3 GUIDE BUNDS

6.3.1 Form in Plan

Guide bund may be given an elliptical shape as in Figure

6.1 since separation of flow can be avoided better with such an
alignment. Alternatively straight parallel guide bunds can also be
adopted if preferable on account of any reasons. Straight parallel
guide bunds may be provided with heads of sufficiently flat
curvature or of composite curve, to avoid separation of flow from
the guide bund head and shank.

In case of submontane rivers and when there are site

constraints, deviation from elliptical shape may be warranted.
Guide bunds shapes conforming to channel configuration and
site conditions with wide splay and along length may then be
necessary.

6.3.2 Length of guide bund

For encouraging axial flow through the bridge, upstream

length of guide bund may be kept 1.0 to 1.5 times the bridge
length and downstream length 0.25 to 0.40 times the bridge length.
Length necessary for keeping worst possible embayment behind
the guide bund sufficiently away from the approach bank is also
required to be ascertained by fitting of sharpest loop as indicated
in Figure 5.2 in Chapter 5. If field data of acute bends in river
when cut offs occurred is not available, It may be assumed that
radius of such a bend can be equal to 0.4 times the radius of
average bends if maximum river discharge is up to 5660 m3/s
and equal to 0.5 times the radius of average bends if the maximum
discharge is more than 5660 m3/s. In case of braided rivers, the
length determined on basis of both the considerations of axial
flow through bridge and keeping away the worst loop may be
found to be insufficient for protection of long approaches.
Additional protection works like spurs, revetments etc. may then
become necessary.

127
 Fig. 6.1 : Elliptical shape of guide bunds

W (CM/SEC.)
0.01 0.02 0.04 0.06 0.1 0.2 0.4 0.6 1.0 2 4 68 10 20 40 60
10 10
8 8
6 6
4 4
0 0
2 2

1 1
0.8 0.8
0.6 0.6
D
0.4 0.4
(MM) T = 320 F 0
60 F
0.2 0.2
0.1 1000 F 0.1
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.1 0.1
0.01 0.02 0.04 0.060.1 0.2 0.4 0.6 1.0 2 4 6 8 10 20 40 60
D = DIAMETER OF SEDIMENT GRAIN (MM.)
W = SETTLING VELOCITY (CM./SEC.)
Fig. 6.2 : Settling velocity of sediment grains of
different sizes

128
 On the other hand if constriction is small on one or both
sides and exposed length of approach is accordingly short, guide
bund of length smaller than the length of the bridge may be found
adequate to ensure safety of the approach from worst loop and
also to guide flow smoothly without separation, obliquity and
with optimum uniform discharge distribution across the bridge.
Shorter length than bridge length for the guide bund is considered
permissible under these conditions.

6.3.3 Curved heads

In case of rivers carrying high discharges of the order of

28300 m3 Is and above, Gale’s recommendations are more
conservative than Spring’s. For rivers with discharge of the order
of 14150 m3/s, radius of curve according to different authors is
nearly same. For discharges smaller than 14150 m3/s radius
can possibly be shorter than suggested by Spring. In the absence
of definitive overall criteria, judgement needs to be used in
selecting suitable radius or radii when the curve is composite. In
case of major bridge projects, aid of hydraulic model studies
may, in addition, be sought in finalising the curve.

6.3.4 Cross section and protection

(i) Top Width

Sufficient top width may be provided for the guide bund

to permit easy transport of construction materials. It should be
in any case not less than 6 m.

(ii) Top Level

Top level of guide bund should allow a minimum free

board of 1 m above affluxed high flood level corresponding to
foundation design discharge. On the rear side of the guide bund,
water level rises on account of ponding by an amount equal to
velocity head. When heavy wave action is expected, free board
should be suitably increased.

129
(iii) Side Slopes

The side slopes depend on the nature of river bed material

of which guide bund core is formed and the height. Ordinarily
side slope of 2: 1 on riverside and 2:1 to 3:1 on the other side
may be found adequate.

(iv) Material

Core of the guide bund section may be constructed with

non-cohesive bed material such as sand and gravel, obtained
from riverside and not from back side. Clay should not be used
for core construction.

(v) Size of pitching stone

‘One man stone’ of 30 cm to 40 cm size weighing from

40 to 70 kg which can be handled by a worker may be used.
Such a stone can withstand an average velocity up to about 3.5
m/s. When average velocity is smaller, upto about 2.5 m/s, burnt
brick pitching on edge laid by hand may suffice. For average
velocity higher than 3.5 m/s, crated stone or concrete blocks
may be used, their size or weight being in accordance with the
relationship
V = 5.7 D1/2
and W = 0.04 V6

wherein V is the average velocity in m/s, D is the diameter in m

of spherical crate of equivalent volume and W is weight in kg of a
crate of non-spherical shape. Suitable allowances are required
to be made to this size and weight as explained further. Crates
may be made out of 8 gauge G.I. wire of 4.064 mm dia with
double knots and closely knit to prevent stones slipping out.

Incident velocity on the stone or crate is always less

than the average velocity on the vertical, the relation between the
two being (6.3)

0.71 Vaverage
Vbottom = ———————————
0.68 Log y/D50 + 0.71

130
wherein y is the flow depth in m and D50 is also in m. Adoption of
average velocity in place of bottom velocity introduces a certain
factor of safety.

The V-D relationship is meant for stones with and without

crates having weight of 2645 kg/m3 or specific gravity of 2.65. If
specific gravity of stones is different, the size can be corrected
using Creager equation.

K = 102.5 K
w
0.672 w _ 62.5
wherein Kw is the size in m with weight of stone of w
(kg/m3) and K is the stone size in m with weight of 2645 kg/m3.

When stone or a crate is placed on a side slope, its

stability is reduced. Relation between permissible velocity for
stability on side slope to that on horizontal bottom is given by (6.3)
1/4
Permissible Vbottom on side slope ⎡ Sin 2 φ ⎤
= 1− =x
Permissible Vbottom on horizontal bottom ⎢⎣ Sin 2θ ⎥⎦
wherein φ is the angle of side slope and θ is the angle of repose
of othe stone. Values of x for various side slopes assuming θ of
40 are given below.

Side Slope 1½:1 2:1 3:1 4:1

Value of x 0.71 0.855 0.935 0.953

Permissible velocity V on horizontal bottom is required

to be multiplied by the factor x to find permissible velocity on
side slope for given size of stone or crate. Alternatively for given
velocity, if size of stone or crate on side slope is to be determined,
V is required to be multiplied by 1/x. The relationship
V = 5.7D1/2 is meant for use when side slope is 2 horizontal to 1
vertical. For side slopes other than 2:1 suitable multiplying factors
have to be adopted. Additional aIlowances in stone or crate size
to account for severity of attack, obliquity of flow, wave action,
higher level of turbulence etc. should be made using Table 5.8 of
Chapter 5 for guide.

131
(vi) Thickness of pitching

Pitching constructed of one man stone weighting 40 to

70 kg is used in Indian practice and no further gradation is allowed.
Stones of this size have been found to be stable in velocities up
to about 3.5 m/s inclusive of all factors such as severity of attack,
slope effect etc. After considering all available data, Inglis
recommended that the thickness T of pitching should be worked
out according to formula T =0.06 Q1/3, wherein Q is discharge in
m3/s and T thickness of pitching in m. This formula is therefore,
proposed for adoption out of various formulae referred in Chapter
5.

Thickness is increased by 25 per cent at head to take

care of severe attack due to exposure and 25 per cent all over
when pitching is dropped through deep water below low water
level. Thickness can be reduced by 150 to 225 mm by using
quarry refuse or burnt bricks as filter. The final thickness so arrived
at should not however exceed the upper limit of 1.3 m.

In following theoretical practice, the stones to be used

in pitching should be well graded is as indicated below.
Size of Stone Percentage of total weight
smaller than given size

3k 100
2k 80
1k 50
0.1 k not to exceed 10

in which k is the diameter of stone that will have the weight

same as 50 per cent size of stones D50 arrived at applying the
formula
V = 5.7 D1/2
D50 = 0.031 V2
W = 0.04 V6
wherein W is in kg/m3 and V and D50 are in m.

132
 Thickness of such graded stone protection should be
2D50 max or 1.5D100 max whichever results in greater thickness.
Thickness should not be less than 1.5 D50 and not less than 31
cm.

The graded stone pitching is placed on a filter of proper

design. Size, gradation and thickness of the filter material should
be such that bank material is not sucked out through voids.
Standard specifications generally require –

D15 of riprap D15 of riprap

————————— < 5 < ————————— < 40
D85 of bank material D15 of bank material

The thickness of filter blanket ranges from 15 to 23 cm

for a single layer and from 10 to 20 cm for individual layers of a
multilayer blanket.

Synthetic mesh nets of woven fabric provide a good

filtering medium and can replace granular filters especially when
laying filter under water becomes difficult.

6.3.5 Launching Apron

(i) Size of stone

Size of apron material may be kept same as for side

slope protection provided thickness is according to Indian practice
advocated by Inglis and others. Otherwise size can be as
recommended by the ISI viz

W = 0.10 V6

(ii) Thickness

Thickness of apron when launched should be 25 to 50

percent more than the thickness of pitching on slope.

(iii) Slope of launched face

Launching slope may be assumed as 2 horizontal to 1

vertical.

133
(iv) Quantity of apron stone

Quantity should be sufficient to cover the scoured face

down to maximum depth of scour obtained using scour factor c
of 2.5 at upstream curved head and 1.5 for the rest of the length
in Lacey formula
2 1/3

D = c x 1.34 ( )
q
f

wherein D is depth of flow below H.F.L. in m, q is discharge

intensity in m3/s per metre width within guide bunds on bridge
section, f is the silt factor and c the scour factor. It has been
observed that the above Lacey formula does not give a good fit
for data of some of the rivers. As an alternative, the Laursen
formula for guide bund contraction can therefore be applied in
the following form to obtain average depth of flow D within guide
bunds.

[ ][ ]
2/3
Width of unconstricted Discharge of
Depth within channel excluding spill channel plus
guide bunds portion spills
————— = —————————— X ———————
Depth of Width within guide bunds Discharge of
unconstricted channel
discharge

wherein

X = 0.59 when (√ )gDS

w =1/2

X = 0.64 when (√ )gDS

w =1

X = 0.69 when (√ )gDS

w =2

In √gDS/w, g is gravitational acceleration in m/s/s, D is

depth of flow in m, S is river slope and w is velocity of fall of
sediment particles of size corresponding to mean diameter of

134
bed material in still water, which can be estimated from curves in
Fig. 6.2.

The scour depth can be arrived at by using scour factors

mentioned before. Bigger of the two scour depths, obtained using
Lacey formula and Laursen procedure should be adopted for apron
computations. At the curved head, launching surface becomes
fan or cone shaped with length of sloping face equal to 2.3 times
the depth of scour below bed level. The length of fanned out
surface is equal to xπr where r is radius of cone at half the depth
of scour below bed level and x is the sweep angle in degrees/180
as shown in figure 5.7 of Chapter 5. The quantity required for
covering the fanned out surface should be provided over the entire
sweep angle.

(v) Laying of apron

Width of apron laid on river bed should be 1.5 D1 where

D1 is estimated scour depth in metres below river bed. Thickness
of apron as laid can be worked out as quantity of apron stone/
width of apron laid and works out to 1.5 T1 where T1 is the thickness
of apron after launching. The thickness at the head should be
increased to 2.25 T1 and checked up to verify whether the quantity
is adequate for fanning out. If not, the thickness should be
increased further as required. Apron should be turned round the
noses on upstream and downstream sides. Fig. 6.3 illustrates
important design features.

6.4 SPURS OR GROYNES

6.4.1 Types of Spur
As explained in 5.4.2 solid spurs are generally used for
deflecting attack of flood flow whereas permeable spurs are
provided primarily for bank protection when flood velocities are
relatively lower and flood flow carries heavy concentration of
suspended sediment.

6.4.2 Location

Any reach undergoing or threatened by erosion should

be determined by field inspection. Previous surveys of bank line,

135
Fig. 6.3 : Apron design features

136
if available, should be superimposed to demarcate the eroding
reach. In bend flow, more vulnerable portion of bank to erosion is
generally at the downstream end of the bend.

6.4.3 Length, inclination, height and spacing for

impermeable spurs

These details should be fixed following criteria given in

Chapter 5.

6.4.4 Cross section and protection of impermeable spurs

The following guidelines are generally in accordance with

Indian Standards 8408-1976.

Top width is generally kept at 3 to 6 m to allow vehicular

transport for construction materials.

Free board of 1.0 to 1.5 m is allowed on upstream side

above anticipated high flood level.

If shank is constructed of sand, side slope can be 2: 1,

If stone is used instead of sand side slope can be steeper at
1 ¼ :1 or 1 ½ :1. Nose slope should, however, be 2:1 irrespective
of whether it is constructed in sand or stone. It is preferable to
construct nose wholly in stone.

Stone pitching should be provided on all exposed slopes

of the spur. Size of stone on sloping face should withstand the
estimated flood velocity. Ordinarily stone of 40 to 70 kg weight
can withstand flow with average velocity up to 3.5 m/s on side
slope of 2:1. For higher velocities, stone crates or concrete blocks
have to be used. In concrete blocks of required size, pattern
grouting may be done which helps to localise damage and adds
stability. In general side slope protection should be on the same
lines as in case of guide bunds.

The thickness of pitching should be designed in the same

way as in the case of guide bunds. The thickness can be reduced
from nose backwards along the length of the shank as given in
Table 6.2.

137
 Table 6.2

Length of spur to
Thickness of be provided with
Location pitching T metres pitching of thick-
or fraction of T ness in Col.2

1 2 3

Nose T Whole

Upstream side T First 31 to 46 m or

of shank scour hole length
whichever is larger.

2/3 T Next 31 to 46 m

Nominal 0.3 m Rest

2/3 T First 31 to 61 m
Downstream side
of shank Nominal or Nil Rest depending on
action of return flow
Graded filter 20 to 30 cm thick satisfying standard design
criteria should be provided at the nose, and over the shank. When
thickness of pitching is kept nominal, thickness of filter can be
reduced to 15 cm or it can be omitted considering likelihood of
parallel flow and its action.
Apron of standard design as in case of guide bunds
should be provided for protection of scoured face down the entire
depth of scour hole.

Size of apron stone or crates or concrete blocks should

be the same as in case of guide bund. Slope of launching apron
may b assumed to be 2: 1.

Scour depth may be found according to the norms given

Indian Standard 8408(6.4) or after experiments conducted by
Mustaq Ahmed (6.5) as in Table 6.3.

138
Table 6.3 : Scour Factors for Apron Design in case of Spurs
(A) According to Indian Standard Specifications
Sr.No. Location Scour Factor
(i) Nose 2.0 to 2.5
(ii) Transition from nose to 1.5
shank and first 30 m to 61 m
on upstream
(iii) Next 31 m to 61 m on upstream 1.0
(iv) Transition from nose to shank 1.0
and first 15 m to 30 m
on downstream

(B) On Basis of Experiments conducted by Mushtaq Ahmed

Sr.No. Location Scour
Factor

(i) Nose
For spur facing D/S at 30o to bank line 1.5
For spur facing D/S at 60 to bank line
o
2.0
For spur at 90 to bank line
o
2.5
For spur facing U/S at 30 to bank line
o
3.0
(ii) Shank
For spur facing D/S at 30o to bank line
On U/S face over length 0.85 L 1.0
On D/S face over length 0.30 L 1.0
For spur facing D/S at 60o to bank line
On U/S face over length 0.80 L 1.0
On D/S face over length 0.40 L 1.0
For spur at 90o to bank line
On U/S face over length 0.70 L 1.0
On D/S face over length 0.50 L 1.0
For spur facing U/S at 30o to bank line
On U/S face over length 0.60 L 1.0
On D/S face over length 0.72 L 1.0

139
Note :-
(a) Scour factor is defined as scoured depth below
H.F.L. divided by Lacey depth below H.F.L.
(b) L is the projected length of an inclined spur given
by actual length x sine of angle with respect to
bank line.
(c) Out of the apron designs worked out according
to A and B above, heavier of the two may be
adopted.

Practice followed in this respect in the U.S.A. is given in

Table 6.4 for comparison.

Table 6.4 : Formulae in Vogue in the U.S.A.

for Prediction of Scour at the Noses of Spurs
S.No. Author Formula

1. Laursen(6.9) L d ⎡⎛ 1 d s 1.7 ⎞ ⎤
= 2.75 s ⎢⎜ . + 1 ⎟ − 1⎥
D D ⎣⎝11.5 D ⎠ ⎦
0.4
2. Liu(6.14) ds ⎡L⎤
= 1.1⎢ ⎥ Fr 0.33
D ⎣D⎦
Recommended by Richardson(6.15)
L
< 25
D
3. For spurs in ds
the Missis- = 4.0Fr 0.33
D
sippi river with
Recommended by Richardson(6.15)
large L/D ratio
L
for > 25
D
For arriving at scour depth with spurs making different
angles with flow direction, multiplying factors are used.
Note :- In the above formulae :
dS is equilibrium scour depth below mean bed level, D is upstream
depth flow, Fr is upstream Froude Number ⎛⎜ V ⎞⎟ and L is
⎜ ⎟
effective or projected spur length. ⎝ gD ⎠

140
 Thickness of apron should be 25 to 50 percent more
than thickness of pitching on slope.

Width of launching apron should be equal to 1.5 times

the depth of scour below river bed. At the nose, additional quantity
of stone should be provided to allow fanning out along the cone
face to scour level as in case of upstream head of a guide bund.

A typical design of spur is illustrated in Fig. 6.4.

6.4.5 Permeable Spurs

A permeable spur may be constructed by driving two or

more rows of bally piles with sufficient penetration below riverbed
level to provide adequate grip during floods. Cross braces and
longitudinal runners or waling pieces may be provided to
strengthen the structure. Stays may be added on either ends so
that differential pressure and flood attack can be withstood. The
structure may be filled with brush wood and weighed down by
stones. For more details on permeable structures, Chapter 7
may be referred.

6.5 APPROACH BANKS AND MARGINAL

EMBANKMENTS

6.5.1 Approach Banks (6.5)

When approach banks are of medium height upto 3 m,

the material may be homogeneous. For bigger heights zoned
section with impervious core and pervious outer sections may
be desirable.

Foundation should provide a stable support for overlaid

bank and should resist harmful percolation. Unconsolidated
material in foundation is liable to settlement and percolation of
water. Cut off trenches backfilled with rolled impervious material,
sheet pile cut off, grouting, clay blanket on upstream side and
inverted filter on downstream side are helpful in minimising harmful
effect of percolation.

Hydraulic gradient should be well within the bank with

141
Fig. 6.4 : Design of impermeable spur

142
adequate cover. As an approximation it may be assumed as a
straight line with slopes of 1: 1 in impervious clay varying to 1: 12
in sandy soil. For ordinary clay the gradient is about 1 in 4. The
cover over hydraulic gradient line should be 0.9 to 1.2 m along
slope and about 0.6 m along ground.

The approach bank section should be stable for

conditions of saturation and sudden drawdown. On the
downstream side worst condition for stability can be when there
is H.F.L. on upstream side and much lower water level on
downstream face. Under all these conditions a factor of safety of
1.3 to 1.5 should at least be ensured.

Top width is guided by consideration of locating railway

line or road way.

Top level for approach bank should be worked out by

adding to the design high flood level allowances for afflux, rise in
water level from bank to guide bund head due to slope, velocity
head, wave wash and free board. For upstream slope, the angle
should be flatter than the underwater angle of repose of the
material. The steepest slope for good soil can be 2:1 for low
embankment of height up to 3 m and 3:1 for high embankments.
If the material is sandy, the upstream slope may be as much as
4:1.

Downstream slope should be guided by hydraulic gradient

and cover. If on account of any reason, removal of material due to
piping is noticed, addition of inverted filter and stone pitching
may be found to be useful to check erosion.

Side slopes and cross section of the approach bank

should also be designed, especially in case of high embankments,
on stability considerations based on engineering properties of
the material forming the bank and foundation and hydraulic
conditions.

Side slope protection should be provided against parallel

flow on the same lines as in case of guide bunds. On the down-
stream side of approach banks, turfing may be provided for
protection against rain and wind. Longitudinal and cross drains
may be considered if rainfall is heavy and embankment high.

143
 When bridge approaches obstruct spill discharge and
guide bunds are not provided, deep scour holes can be formed at
abutments. Estimation of scour depth and protection at the
abutments for slope and for toe in the form of apron may be
similar to that in case of spurs.

6.5.2 Marginal Embankments

Design aspects of marginal embankments are similar

to those of approach banks. The aspects needing special
attention are dealt with below.

Distance of marginal embankment from the river bank

should neither be too short nor too long. If the embankment is
too near the river, velocity along the bank can become high.
Normally this velocity is not allowed to be more than 0.9 to 1.2
m/s. On the other hand if the embankment is far too away, the
very purpose of providing the embankment can be defeated.
Current Indian practice is to roughly maintain a distance of 3
times the width given by Lacey formula in between the flood
embankments on either banks of the river, though this thumb
rule may not be applicable in case of all alluvial rivers. The flood
embankment is carried on a ridge or high bank. In alluvial rivers a
ridge or high ground is often formed near the bank line due to
deposition of suspended sediment when river overflows the bank.
Protrusions and sudden changes in alignment forming kinks are
as far as possible avoided. Structures like bridges and barrages
obstruct free downstream movement of meanders. Immediately
upstream of such structures, the meander belt, therefore, widens
and becomes almost double. While fixing location of marginal
banks, this field experience needs to be kept in view.

Height of embankments may be designed by allowing

sufficient free board above design high flood level obtained by
routing design flood hydrograph.

Since velocities along river side slope of the marginal

banks are expected to be low, no slope protection may ordinarily
be required and turfing on both side slopes may suffice for
protection against rain and wind. In special circumstances, if
river attack warrants slope protection, it can be similar to that
suggested in case of guide bunds.
144
6.6 BANK REVETMENT

6.6.1 Side Slope

Angle of bank should be flatter than angle of repose of

pitching material. Angle of repose of sand is 2:1 while that of
stone is about 1:1. Angle of river bank formed of cohesive soil
can however be steeper than 1:1 in which case it is required to
be graded and made flatter than 1: 1. Flatter the side slope with
respect to angle of repose of pavement material, the size and
weight of pitching stone can be reduced.

6.6.2 Velocity in the eroding bend along concave bank

If the average velocity in the river in a straight reach is

known but not the velocity along eroding concave bank in the
bend, it can be estimated using the relationship (6.3)
0.5
Tbend V 2 bend ⎡r⎤
= 2 = 3.05⎢ ⎥
Tstraight V straight ⎣w⎦

wherein Tbend is maximum boundary shear in kg/cm2 as affected

by bend, T straight is average boundary shear in kg/cm2 in a straight
reach, r is center line radius of bend, and w is water surface
width in m at upstream end of the bend.

If average velocity in the bend section is known,

corresponding velocity along eroding bank can be roughly

[ ]
estimated also on basis of flow depths using the expression
2/3
V at eroding bank Flow depth at eroding bank
—————————— = —————————————
V of cross section Average flow depth of the cross
mean
section

When measured flow depth along eroding bank is not available,

the following guide lines given by Lacey(6.6) for estimation of the
depth may be found useful.

145
 Pattern of channel
i) Straight reach
ii) Moderate Bend
iii) Severe bend
iv) Right angled bend
Increase in flow
depth below
High Flood Level
…1.25 x normal flood depth
… 1.50 x normal flood depth
… 1.75 x normal flood depth
… 2.0 x normal flood depth

6.6.3 Side slope and apron protection

Design of protection for side slope and of apron may be

on similar lines as for guide bunds.

6.7 ARTIFICIAL CUT OFFS

Alignment of a cut off is chosen so that approach and

exit ends remain more or less tangential to the river course. A
cut off can be made either by open excavation or by means of a
dredger. Since considerable amount of work is usually involved
below low water level making a cut becomes easier with a dredger.
Normally only a small pilot cut of about 10 per cent capacity of
the channel discharge is made where the river cut off is desired.
The river is relied upon to open out the pilot cut and develop it
once the water starts flowing. When arc to chord ratio is about 5
or more, development of a cut off may be expected to be rapid. A
deep and narrow cut may be preferred to a shallow and wide
section, since side erosion is easier and faster than bed scour.
Velocities in deeper cut are also more.

In designing an artificial pilot cut, its progressive

development and final adjustment of the river are required to be
visualised. When cut off is made across the neck of the bend,
length is shortened so that with the available fall, slope steepens
increasing the velocities. These increased velocities open out
the cut rapidly and the river gets diverted on the new cut off
channel. With increased velocities, bed scour and degradation
occurs upstream of the cut off. The scoured material is deposited
on the downstream side and finally equilibrium bed and water
surface slopes are established which can be a little steeper than

146
the original. Thus in effect river attains a somewhat steeper slope
and generates higher velocities. Bank protection is, therefore,
usually needed at and in the vicinity of a cut off to prevent side
erosion. Otherwise new bends may form in the wake of a cut off
as was experienced in the Mississippi River(6.7). Superimposition
of the expected longitudinal section with cut off over the existing
section permits a rough assessment of possible degradation and
aggradation. Such an attempt should be made to estimate the
final bed and water levels and lengths likely to be affected by the
cut off.

6.8 BRIDGE PIER PROTECTION

6.8.1 Deep piers

(i) Scour depth and grip length

Present practice is to sink bridge piers sufficiently deep

to provide a safe grip length even after full scour occurs around
them at the foundation design flood stage.

Spring suggested minimum depth of pier foundations

based on anticipated scour as shown in Fig. 6.5.

According to this figure the grip below scoured bed level

is half the height from scoured bed level to H. F. L. for 30 m
height, less than half the height for bigger depth and more than
half for lesser depth of flow.

The provision in IRS and IRC codes is that minimum

grip should be 2/3 the flow depth in the bridge or 1/4 the total
height of pier from bottom to design H. F. L. or 1/3 the flow depth
below design H. F. L. after pier scour has occurred. When this
grip is ensured by sinking deep wells, no protection against scour
is required to be given.

Estimation of scour depth is thus an important aspect

of pier design. According to Indian Railway practice based on
studies made by Inglis (6.7) pier scour is assumed to be double
the normal depth of flow given by Lacey formula

D = 0.473 (Q/f)1/3

147
 123 M

{
0 DE A 0
EP 3.05 GREATEST DEPTH OF
ES 6.09 PIER IN WATER AS
TA 9.14
SC 12.19 MEASURED BELOW
OA REPRESENTS HIGHEST ER 15.24 HIGHEST FLOOD.

DE
KNOWN FLOOD LEVEL TA
INA 18.29

PT
BL 21.34

HO
OB REPRESENTS HIGHEST ES 24.38
CO 27.43

FB
{
KNOWN FLOOD LEVEL UR 30.48

OT
33.53 LEAST DEPTH OF

TO
OB REPRESENTS DEPTH 36.58 PIER IN THE SOIL

M
39.62
TO WHICH FOUNDATIONS OR BELOW GREATEST

OF
B 42.67
SHOULD BE SUNK. PROBABLE SCOUR.

FO
45.72

UN
48.77
TI

DA
51.82
ON 54.86
S 57.91
C 60.96
NOTE :

THE DIAGRAM APPLIES ONLY TO SANDY BOTTOM. IF THE RIVER BED

IS SOFT SLUSH A GREATER DEPTH IS NECESSARY. PIERS ARE ALWAYS
PRESUMED TO HAVE ENOUGH STONE AROUND THEM TO PREVENT LOCAL
PIER-FORMED SWIRLS FROM SCOURING POT-HOLES AT PIER BASE.

Fig. 6.5 : Depth of bridge piers according to spring

Fig. 6.6 : Relation between pier size and scour depth

given by Laursen

148
 The pier scour in IRC code is assumed to be double the
normal depth of flow given by Lacey formula

D = 1.34 (q2/f)1/3

When bridge is provided with constricted waterway, bigger

of the two scour estimates should be used in design of pier
foundations. In the above two Lacey formulae, q is the intensity
of discharge in m3/s per meter width within abutments, D is depth
of flow in m nearly equal to hydraulic mean radius in case of wide
rivers, Q is the design discharge in m3/sec and f the silt factor
given as 1.76 m1/2, m being the weighted mean diameter in mm
of the bed material. The weighted mean diameter is obtained by
giving weightage with respect to percentage fractions of difference
sizes.

If m has nearly the same value over the entire scour

depth, working out silt factor presents no difficulty. It is however
possible that size of bed material may change at various depths
from riverbed to scoured bed level. Silt factors are then worked
out for each strata having nearly uniform sized material and scour
depth determined adopting each f value starting from rivers bed
downwards. If size of bed material is coarser at lower levels, final
scour depth will work out to be less than given corresponding to
riverbed material. If material happens to be finer at lower depths,
the scour depth will work out to be more.

Procedure to be followed in working out scour depth when

bed material comprises different sizes at various elevations is
given below:

Estimation of Scour Level in River Bed with Varying Sizes

of Bed Material at Different Depths.

Design Data

(i) Foundation design discharge = 1500 cumecs

(ii) Corresponding H.F.L. = 107 m

(iii) River bed level at bridge site = 100 m

(iv) Weighted mean diameter of bed material

at different depths.
149
Alternative (A) –

From RL 100 m to 98 m – size of bed material 0.32 mm

From RL 98 m to 97.2 m – 0.50 mm

From RL 97.2 m to 90 m – 0.70 mm

Alternative (B) –

From RL 100 m to 98 m – 0.50 mm

From RL 98 m to 96.2 m – 0.32 mm

From RL 96.2 m to 90.0 m – 0.20 mm

It is required to estimate the scoured bed level at a bridge

pier adopting scour factor 2 in the Lacey formula

D = 0.473 (Q/f)1/3

The stepwise procedure is. given in Table A for alternative

(A) and Table B for alternative (B) respectively.

Estimation of scour depth using Lacey formula is an

empirical procedure which ignores size of obstruction. No
additional allowance for non-uniform discharge distribution, flood
scour, bed form etc. is normally made.

Laursen developed a semi-theoretical approach for

estimation of pier scour which accounted for the size of the pier.
By using Manning flow formula and his own sediment transport
function, Laursen first obtained scour depth ds below river bed in
case of a long contraction. The scour at abutment was then
assumed to be r times that in a long contraction. Value of r was
found to be 11.5 by model tests. To obtain pier scour, width of a
rectangular pier b (or diameter in case of a circular pier) was
considered equal to twice the effective length of the abutment.
The Laursen relationship so obtained was the following shown
graphically in Fig. 6.6.

150
 Table 'A' : Scour level computations for bed material sizes given in Alternative (A)

S. Elevation Value Value of 'f' D Lacey Scour Level Whether Remarks Estimated
No. of 'm' f = 1.76m½ D=0.473 Q 1/3 HFL-2D Lacey estimated scour scour level
( )
f level is higher to be finally
than, equal to or adopted
lower than level
corresponding to
'm'

1 2 3 4 5 6 7 8 9

1. From 100 to 98 m 0.32 mm 1.00 5.45 m 107.00 - 10.90 Lower Scour level

151
= 96.10 m will be below
RL 98 m

2. From 98 m to 97.2 m 0.50 mm 1.25 5.00 m 107.00 - 10.00 Nearly equal Scour will Scoured
= 97.00 m continue to level
RL 97.2 m will be
RL 97.2 m
3. From 97.2 m to 90 m 0.70 mm 1.47 4.77 m 107.00 - 9.54 Higher Scour will
= 97.46 m not proceed
below
RL 97.2 m
 Table 'B' : Scour level computations for bed material sizes given in Alternative (B)

1 2 3 4 5 6 7 8 9

1. From 100 to 98 m 0.50 mm 1.25 5.00 m 107.00 - 10.00 Lower Scour level
= 97.00 m will be lower
than
RL 98.00 m

2. From 98 m to 96.2 m 0.32 mm 1.00 5.45 m 107.00 - 10.90 Nearly equal Scour will Scoured
= 96.10 m continue to level
RL 96.10 m will be
RL 96.00 m

152
3. From 96.2 m to 90 m 0.20 mm 0.835 5.5 m 107.00 - 11.00 Higher Scour will
= 96.00 m cease at
RL 96.0 m
 d s ⎡⎛ 1 d s ⎞ ⎤
1.7
b
= 5.5 ⎢⎜ + 1⎟ − 1⎥
D D ⎣⎝11.5 D ⎠ ⎦

This relationship was found to compare well with a similar

relation evolved by Laursen on basis of model test results
independently. According to the latter when depth of flow was
more than 5 times the pier diameter, the scour depth below river
bed approximated 2.75 times the pier diameter. Work of Shen(6.16),
Tarapore(6.17), Larras (6.10) and Breusers however indicates that
limiting equilibrium scour depth can be less than that given by
Laursen procedure. Tarapore found that when depth of flow was
more than two times the diameter of a cylindrical pier, depth of
scour below river bed approximated 1.4 times the pier diameter.
According to Breusers relationship based on model data gives
scour depth equal to 1.40 times pier diameter. For equilibrium
scour with continuous sediment movement, criteria by Larras
and Breusers were found by Shen to give an envelope for all
known data. In view of the variation in estimated scour depth in
terms of pier diameter according to various formulae, it appears
safer to adopt bigger of the estimated values as obtained by
Laursen curves. It is necessary to make further allowance to this
scour depth to account for lowering of river bed due to constriction
by guide bunds, non-uniform discharge distribution, flood scour,
bed form, etc.

Out of the two methods available for scour estimation,

one based on discharge in case of Lacey formula and the other
based on pier size as in case of Laursen relationship, superiority
of one over the other has not been convincingly proved. It is,
therefore, considered advisable to work out estimated scour depth
by both the procedures and adopt higher of two values.

(ii) Flood Scour

River bed surveys are usually made when river flows are
low. During floods river bed may be lowered due to scour. For
estimation of flood scour, riverbed is required to be gauged on
the bridge section line at various flood stages and flood scour
estimated by extrapolation of this data of intermediate floods to
design flood stage.

153
(iii) Scour due to constriction

Lowering of river bed level on account of constriction of

waterway by guide bunds has to be considered separately when
scour depth is assessed on basis of size of the bridge pier. For
estimation of constriction scour, three procedures are available
which are due to Laursen, Latischenkov and Lacey (6.12). All these
being very similar, Lacey formula can be used to give
0.67
D 2 ⎡ W1 ⎤
=
D1 ⎢⎣ W2 ⎥⎦

wherein D1 and W1 are depth of flow and width in m in unconstricted

section and D2 and W2 are corresponding values on the bridge
section within guide bunds.

(iv) Effect of bed forms

Lowering of bed level due to movement of big sized bed

form is also relevant in assessment of pier scour level. Sand
waves of 6 to 9 m height have been observed to form and move
on bed of big rivers like Brahmaputra. When a trough passes
across the pier, the depth would accordingly increase. In the
absence of reliable predictors for height of bed form, field
observations are needed to account for this factor.

(v) Pier scour in clayey strata

The concept of clear water scour was developed by

Laursen(6.18) for application to scour at bridges on overbank spill
portion where general bed load transport is absent. This concept
can be considered valid also in case of scour in cohesive soils.
The clear water relationship was obtained in case of abutment
scour by Laursen as
⎡⎛ L d ⎞
7/6
⎤
⎢⎜ s
+ 1⎟ ⎥
L ds ⎢⎝ r y0 ⎠ ⎥
= 2.75 1/ 2 − 1
y0 ⎢
y 0 (To / Tc) ⎥
⎢ ⎥
⎣ ⎦
154
wherein L is effective length of the abument in m obstructing
flow, yo is the depth in m of approaching flow, ds is depth of scour
in m below bed level, r is a constant being the ratio of scour
depth at abutment to that in a long contraction and was found to
have value of 12, To is shear stress in kg/cm2 on bed in the
approach and Tc is critical tractive force in kg/cm2 for the material
eroded from the scour hole at the abutment. Scour at bridge
piers was evaluated by substituting b = 2 L where b is width. of
pier or diameter in case of cylindrical piers as was done for
evaluating scour with sediment transporting flow as in Fig. 6.5.
Value of To can be roughly obtained as γYoS wherein γ is specific
weight of water in kg/m3 and S is slope of approaching flow.
Value of Tc for cohesive materials can be adopted from literature.
Some data in this respect is given in Tables 6.5 and 6.6.

Table 6.5 : Etcheverry’s Maximum Allowable

Tractive Forces Given by Lane

Material Tractive Force TC in kg/m2

Sandy loam 0.35 – 0.40

Average loam, Alluvial soil 0.40 – 0.50
Firm loam, clay loam 0.50 – 0.77
Stiff clay soil 1.35 – 2.12
Conglomerate, soft slate, tough 3.10 – 5.6
hardpan, soft sedimentary rock

Table 6.6 : USBR Limiting Tractive Forces in kg/cm2

Compactness of bed
Description
Loose Fairly Compact Very
Compact Compact
Sandy clays (sand 0.20 0.77 1.6 3.1
content less than 50%)
Heavy clayey soils 0.15 0.69 1.5 2.75
Clays 0.12 0.61 1.38 2.60
Lean clayey soils 0.10 0.48 1.05 1.74

155
 Laursen opined that some of the assumptions made in
evolving the above relationship were rather bold. In the absence
of any better approach, the Laursen relationship is, however,
considered to provide at least some indication of the extent of
scour in cohesive materials. Considerable judgement and caution
are accordingly required to be exercised while extending the
concept of clear water scour to scour in clayey bed as observed
in Reference 6.19.

6.8.2 Shallow Piers

In case of some of the existing bridges, piers may be shallow

and grip may be found to be inadequate. Protection round piers
is then required to be given for their safety.

Minimum grip essential for safety is worked out as 1/4 the

total height of existing pier from bottom to design H.F.L. At this
level bed scour is required to be arrested by providing stone apron
with adequate quantity to fully cover scoured face. Launching
slope in sandy bedded rivers is assumed as 2 horizontal to 1
vertical. Size and thickness are assumed same as in case of
apron of the guide bunds. If one man stone is found to be of
insufficient weight to be stable, concrete blocks or stone crates
are required to be used. Placing stone round piers at unnecessary
higher level is inadvisable as it is liable to be displaced and washed
down more easily. At the same time it is not a safe and feasible
proposition to wait till river bed scours to just the safe level of 1/
4 height to ensure minimum grip length. Dumping of stone has,
therefore, to be resorted to a little earlier before the lowest safe
level is reached.

Even after pitching is placed round the piers, sufficient

portion of the river bed in between the adjacent piers should be
left exposed so that free scour can occur. If this aspect is lost
sight of, pitching of adjacent piers may practically join each other
and pier protection will act as a weir. Scour downstream under
such conditions has been experienced to be bigger, up to 4 times
the flow depth given by Lacey formula which is dangerous.

In case of existing bridges with short spans and shallow

piers, it may not be possible to avoid weir effect resulting on

156
account of pitching round the piers. Sufficient quantity of stone
needs to be provided under such conditions to cater for full
estimated scour depth on downstream side.

Similarly in case of existing bridges provided with drop

walls, it may not be possible to avoid formation of hydraulic jump
at certain flood stages. Provision of adequate quantity of stone
on downstream side to cover scoured face by launching becomes
necessary also under these conditions.

Where conditions permit, instead of dumping stone round

piers, concrete blocks of adequate weight can be laid over filter
covering entire area liable to scour. Protection to piers of road
bridge on Godavari river near Toka in Maharashtra was of this
type (6.13). These blocks should· be laid at a level sufficiently deep
allowing safe grip. Filter below blocks is essential to avoid sucking
out of river bed material through gaps in blocks.

6.9 DESIGN FOR PROTECTION WORKS :

INSTRUCTIONS AS CONTAINED IN IRBM PARA 818

Minor Bridges:

Most of minor bridges are on open foundation. They have to be

properly protected by a well designed flooring system. This will
include floor, curtain and drop wall. Length of floor and depth
of drop wall will be on the basis of scour depth. This can be
determined either by local observation or by using empirical value
of DLacey based on design discharge. Depth of drop wall s
hould be 1.25times DLacey. Floor should cover the entire
width and length of abutment including wing wall. The slope of
floor should match the bed slope and also the top of drop wall
should match the slope. It is essential to do proper protection of
the box culvert which relies on uniform ground support for its
designed structural behavior. If the underside is scoured, the
box culvert gets unevenly supported. For this purpose, properly
designed floor system as described above should be provided.
Sometimes, instead of splayed wing wall, straight return wall
is provided particularly on high bank or in case of a box, another
box is provided to function like a wing wall. Similar protection
work is called for in such cases.

157
Major/Important Bridges:

As far as bridges on open foundations are concerned, it is

generally on rocky/in-erodible bed and not requiring any particular
protection. In other cases, flooring with drop walls as in minor
bridges may have to be provided. Since well and pile
foundationsare designed for the scour, hence no protection
is necessary even in case of a local scour.However, bridge
may need a well designed guide bund with proper protection on
the approach embankments.

River training works through model studies:

In case of large alluvial river, where training/ protection

works involve a heavy financial outlay, model studiesshould
be resorted to, to arrive at the most economical and effective
solution

6.10 DATA REQUIREMENT

Important data required to be collected for design of

training works is indicated below.

6.10.1 Survey data

(i) For bridge design and training works, river plan extending
at least one meander length both on upstream and downstream
showing low water channel, high flood channel boundaries and
extent of overbank flooding with corresponding discharges and
water levels.

Cross sections at bridge site and at upstream/ down-

stream boundaries of the reach with longitudinal section along
the deep channel.

Layout of guide bunds, alignment of approach banks,

ascertaining the necessity of marginal embankments and
protection works, etc. will need such a plan and sections.

(ii) River Plan showing past available courses superimposed

to examine tendency for river meanders to move downstream, to

158
form, cut offs or to develop nodal points. Stability of the river in
plan can be examined with the help of such plan.

(iii) Specific discharge-gauge curves for a sufficiently long

period in respect of both low flow as well as high flood to study
stability against aggradation and degradation.

(iv) Flood cross sections on bridge alignment to assess

flood scour.

6.10.2 Hydraulic Data

i) Design discharge figures for water way and foundation

along with corresponding water levels.

ii) Gauge- discharge or Rating curve.

iii) Normal year and design flood hydrographs.

iv) Frequency curve for annual peak discharges.

6.10.3 Sediment data

(i) Bed material samples across the bridge alignment at

surface and at various depths down to expected deepest scoured
bed level so as to determine mean diameter and silt factor at
various depths on basis of mechanical analysis. In case, bed
material is cohesive, engineering properties like cohesion, angle
of friction, Atterberg limits etc.

(ii) Suspended and bed load concentration at various flood

stages along with percentage of wash load in suspended load.

(iii) Bank material samples taken during falling floods to

determine nature of bank at various elevations. Samples should
be analysed and results presented as in case of bed material.

159
REFERENCES

6.1 ‘Report of the Committee of Engineers,’ Government of

India, Ministry of Railways, October 1959.

6.2 ‘Estimation of design flood peak – A method based on

unit hydrograph principle’. Flood Estimation Directorate,
Central Water Commission, Government of India, New
Delhi, February 1973.

6.3 ‘Hydraulic design of flood control channel’. Engineering

Manual I, July ] 970, U.S.Arrny Corps of Engineers.

6.4 ‘Indian Standards Institution, IS : 8408-1976, Indian

Standard Criteria for river training works for barrages and
weirs in alluvium’, June 1977.

6.5 ‘Embankment Manual’ Central Water and Power

Commission, New Delhi, Series No I, 1956.

6.6 Lancey G,’ Stable channels in alluvium.’ Paper No 4736,

Minutes of Proceedings of the Institution of Civil
Engineering, London, U.K. Volume 229, 1930.

6.7 Manual of ‘River Behaviour, Control and Training,’

Publication No. 60 (Revised), Central Board of Irrigation
and Power, New Delhi, September 1971.

6.8 Fenton J.M.’ Proposed construction of a combined road

and rail bridge over the river Ganges at Patna or
Mokamehgh at Engineering Report’, East Indian Railway
Project 101, Calcutta, 28th April, 1947.

6.9 Laursen, E.M.’, Scour at Bridge Crossings’, American

Society of Civil Engineers, Transactions, Vol. 127,
Part- I, 1962.

6.10 Larras J., ‘Maximum depth of erosion in shifting beds

around river piles’, Annales des points et chaussees,
Vol. 133, No.4.

160
6.11 Preusers, H.N.C.’Scour around drilling platforms,’
Bulletin, Hydraulic Research 1964 and 1965, International
Association for Hydraulic Research, Volume 19.

6.12 Chitale, S.V.,’River formulae,’ Irrigation and Power

Journal Central Board of Irrigation and Power, New Delhi,
India, April 1980.

6.13 Merani N.V., ‘Repairs and rehabilitation on some bridges

in Maharashtra State,’ Paper 314, Journal of Indian Roads
Congress, Volume 38.2, October 1977.

6.14 Liu, M.K.Chang ,F. M. and Skinner, M.M., Effect of bridge

constricztions on scour and backwater,’ 1961. Colorado
State University, Colorado, U.S.A.

6.15 Richardson, E.V., Karaki, S., Mahmood, K., Simons,

D.R, and Stevens, M.A., ‘Highways in the river
environment, Hydraulic and environmental design
considerations,’ 1874, Civil Engineering Dept., Colorado
State University; Fort Collins, Colorado, U.S.A.

6.16 Shen, H.W., Schneider V.R., and Koraki S., ‘Local scour
arourid bridge piers’, Joumal of Hydraulics Division
A.S.C.E., Nov. 1969. oages 1919-1940.

6.17 Tarapore, Z.S., ‘A theoretical and experimental

determination of the erosion pattern around obstructions
placed in an alluvial channel with particular reference to
the vertical circular cylinders and piers,’ Thesis presented
to the University of Minnesota in 1962 in partial fulfilment
of the requirement for the Degree of Philosophy.

6.18 Laursen, E.M., ‘An analysis of relief bridge scour’,

Journal of Hydraulics Division, ASCE, 1963., pages 93-
118.

6.19 ‘Scour at bridge waterways’, NCHRP Synthesis of High

way practice 5, Highway Research Board Publication,
U.S.A. 1970.

161
 Chapter 7

PLANNING AND DESIGNING OF PERMEABLE

STRUCTURES

Planning and designing of permeable structures is a new

topic for the bridge engineers. Ease of construction, small cost
of construction material, limited skills and experience required
for the execution of the structures at site, speedy works even
during emergencies, quick results, etc are the important benefits.
Various topics involved for the work are discussed below in detail.

Use of permeable screens / spurs for anti-erosion work

are traditional and are commonly followed in Ganga - Brahmaputra
basin. In case of sediment-laden streams, it helps to induce
siltation along the bank resulting in shifting of river channel away
from the anti-erosion works. These methods are easy for
fabrication on the nearby ground close to the site, need no
especially skilled labour and can be constructed with speed. As
only locally available material is used, these methods have been
found very handy in the anti-erosion works during emergencies
in floods. These methods also work as long-term measures in
the areas where good quality stones are costly and / or are not
available.

Essentially, only a dampening action on the velocity of

flow is achieved by a permeable structure, distinguished from
the deflecting or repelling action of an impermeable structure.
The sediment transporting capacity of a flow is highly sensitive
to the velocity. The theoretical considerations have shown that
the weight of sediment particles is proportional to the sixth power
of velocity. Therefore, the dampening of velocity results in
deposition of coarser particles in the downstream direction.

The purpose, overall behaviour and layout of the

permeable structures can be compared to those of submersible
bund, spur and bank revetment. Permeable structures are cost
effective alternative to the river training or anti-erosion works with
impermeable stone spurs.

162
7.1 FUNCTION

Permeable structures can be used either independently

or with a support of other impermeable stone structures or river
training and bank protection measures. Depending upon the
purpose to serve, the permeable structures are constructed in
transverse or parallel to the direction of flow. Permeable structures
serve one or more of the following functions.

a. Training the river along a desired course.

b. Reducing the intensity of flow at the point of river attack.

c. Creating a slack flow to induce siltation in the vicinity of

the permeable structures and in the downstream reach.

d. Providing protection to the bank by dampening the

velocity of flow along the bank.

7.2 CLASSIFICATION

The permeable structures can be classified as follows

a. According to function served, namely, diverting and

dampening, sedimenting.

b. According to the method and material of construction,

namely, bally, bamboo, tree and willow structures.

c. According to the conditions encountered, namely,

submerged and non-submerged.

d. According to the type of structure provided, namely, spur

type, screen type or dampeners (revetment) type.

7.3 PERMEABLE STRUCTURES

Different types of permeable structures made in the river

channel to achieve the desired river training and bank protection
works, viz, screens, spurs, dampeners, etc.

163
 The permeable structures are made up of different types
of smaller units called as elements. Many elements, made up of
bamboos, ballies, RCC poles, etc are arranged in specific pattern
and linked together to form a permeable structure.

7.3.1 ELEMENTS

Different types of elements are used for making

permeable structures. The dimensions specified for the material
are according to the sizes readily and commercially available in
the market. Therefore, variations in the dimensions, depending
up on those available in the market is made in the design.

a. Porcupines - Porcupines are made up of bamboos / ballies,

have cubical shaped box at the central portion with their legs
extending in all directions. The overall size is 2 m to 3 m. The
central box is filled with stones for stability of individual unit during
floods. Fig 7.1 shows a sketch of bamboo porcupine. Photo 7.1(A)
shows a photograph of bamboo porcupine ready for laying at
site.

Fig. 7.1 : Elevation of Typical Porcupine

164
 Photo 7.1 (A) : Typical Bamboo Porcupine

b. Cribs - This is a pyramid type of structure made up of bamboos

/ ballies with a box at the bottom for holding stones for stability
during floods. Size of the box is generally square of size 2 m to
2.5 m at the bottom. Total height of the structure is 3 m to 4 m.
Fig 7.2 shows a sketch of typical bamboo crib.
Photo 7.2(A) shows a photograph of bamboo crib constructed at
site.

Fig. 7.2 : Sketch of typical crib

165
Photo 7.2 (A) Bamboo crib constructed at site

166
c. Bally frames - Permeable bally structures are made up
of main skeleton of large bamboos or ballies. Cross ballies are
used for stability of the structure. Photo 7.3 shows bally frame
structure constructed at site, projecting from the bank into the
river.

Photo 7.3 Bamboo structure projecting from the bank

into the river

d. Tree branches - Branches of trees or trees of short

height are hanged from a wire rope, duly weighted with stones
and are aligned as a spur projecting into the river. The wire rope
is duly anchored on the bank and in the riverbed. Fig. 7.4 shows
schematic sketch of permeable structure constructed with the
help of tree branches.

Fig. 7.4 Sketch of a typical permeable tree structure

167
7.3.2 Construction material

The main criteria for the selection of suitability of the

material are (i) the cost and (ii) easy / local availability. Other
aspects are of secondary importance. Locally available material
like bamboos, ballies, brushwood, willows, bricks etc is mainly
used for the construction of permeable structures. GI wire, GI
wire mesh, wire ropes, nails etc are the other important but
commercially available material used for the structures.

Standard commercially available bamboos of girth 20

cm to 30 cm are used for the porcupines and cribs. Smaller girth
of 20 cm to 25 cm is used for bracings.

Standard commercially available ballies of girth 15 cm

to 25 cm are used for the bally structures. Normally, the larger
girth of 20 cm to 25 cm is used for the main members, whereas,
the smaller of 15 cm to 20 cm is used for bracings

Four to 5 strands of 4 mm GI wire or wire ropes of 2 mm

size containing 3-4 strands are used for interconnecting
porcupines, cribs, and anchor them to the ground.

Bally driven into the ground up to a depth of 2 m are

treated as anchor. Concrete anchors have an anchor rod of size
32-36 mm, well embedded in concrete cube. Wire crates anchors
are of size 1.5 m X 1.5 m X 1.5 m made up of thick wires and
filled with stones or bricks. A concrete block is casted with bolt
and is included in the wire crate anchor. In case of emergencies,
tie wires are joined directly to the wires of the crates.

7.4 DESIGN CONCEPTS

Dampening of velocities is achieved by the use of

permeable structures. If the flow is sediment laden, siltation is
induced in the slack flow region and the channel is shifted away
from the protected reach.

Different aspects of river training and bank protection

works followed for the permeable structures are similar to those
followed for impermeable stone structures. For example, the
criteria for location, orientation, length and number as followed

168
for impermeable stone spurs are followed for permeable spurs
also.

The porcupines, cribs and bally structures are

multipurpose elements used for all types of permeable structures.
Some limitations are however imposed due to inherent weakness
of the structural material and elements.

7.4.1 Selection of elements

The structural elements commonly used are the

porcupines, cribs, bally frames, tree branches and willows. A
suitable combination of the structure and the elements is made
for the design of protection works. Following points are kept in
view while selecting the elements.

The material like trees, bushes and willows are used for
construction of spurs and dampeners, particularly during flood
emergencies.

Permeable structures are usually designed as

submersible, whereas, bally structures are generally designed
as submersible or non-submersible.

In case of shallow water flows and up to a maximum

depth of 3m to 4m, porcupines are used for both spurs and
screens. For maximum depths of flow from 4 m to 6m, cribs are
preferred. For the depths beyond these limits, bally spurs are
preferred. In practice, it is observed that porcupine spurs and
cribs have effectively protected depths twice those specified
above.

7.4.2 Layout in plan

Permeable structures commonly used are the screens,

spurs and dampeners.

Spurs made up of 3 to 4 rows of porcupines or 4 to 6

rows of cribs. Fig 7.5 shows schematic sketch of a typical
permeable spur. On a straight reach, permeable spurs are spaced
at 3 to 4 times its length. On a curved channel, depending up on
the obliquity of flow, the spurs are spaced at 2 to 3 times the

169
length. Projection of the spurs into the river channel is normally
11 % to 15 %. Three spurs are provided for a specific reach to be
protected. A single permeable spur is generally not found effective.
Alignment of spurs is kept pointing towards upstream with
reference to the flow.

Fig. 7.5 Sketch of typical layout of Porcupine Spur

170
 For depth of flow up to 3 m, two rows of porcupines are
laid along the bank on either side of the toe as dampners. For
more depth, numbers of rows are increased.

Permeable screens are used for choking the secondary

channels. Four to 6 rows of porcupines or 6 to 9 rows of cribs are
used in a permeable screen. One screen is normally provided at
the entrance of the bypass or secondary channel. The second
screen is provided at a distance of 1 to 1.5 times width of the
screen and are extended on both the banks for a length one third
of the channel width. Fig 7.6 shows a typical layout for inducing
sedimentation in a secondary channel.

Fig 7.6 : Porcupine screens for inducing sedimentation

7.5 STABILITY OF STRUCTURES

Due to inherent weakness of the elements, the counter

weights are provided (i) in the central box of the porcupines or (ii)
in the bottom tray of the cribs Due care is necessary to tie the
weights to the main body of the elements.

171
 The elements are tied to each other by wire ropes. The
tie ropes are duly anchored to the bank and at the nose with the
help of suitable anchor or anchor blocks. Intermediate anchors
are also provided at an interval of 15 m to 20 m along the length
of the structure on the upstream side.

No bed protection is needed for the structures made up

of porcupines and cribs. Sinking of these structures in to riverbed
is a welcome feature, which adds up to the stability during floods
resulting in better performance.

7.6 PROTECTION TO THE STRUCTURES

Normally, no additional protection is provided to the

permeable structures. In case of estimated higher velocities along
the bank, additional rows of porcupines are provided along the
bank on both sides of permeable spur. Additional bed protection
or protection at the nose can be provided to bally spurs. If so
desired, the size of stones, size of aprons etc can be designed
as per IS 8408. Temporary protection of burnt bricks in wire crates
can be provided to the bally structures or to the nose of permeable
spurs.

7.7 MISCELLANEOUS

In order to divert the flow and reduce pressure on the

protection works, wherever feasible, pilot channels can be
provided in addition to the river training works constructed with
permeable spurs / screens.

In case of high velocity flows, implementation of only

permeable structures is not favoured. However, use of permeable
spurs in between the reach of two solid stone spurs is more
effective.

Care is necessary to see that the size of the stones /

bricks (the minimum dimension) is larger than the maximum
size of openings provided in the crates of GI wire / nylon rope
mesh.

As alternative material, instead of bamboo porcupines,

RCC poles have been successfully used as anti-erosion measure

172
and sedimentation device. The RCC porcupines are able to
withstand higher velocities. They also have longer life than the
bamboo structure. Fig 7.7(A) shows a sketch of typical RCC
porcupine. Photo 7.7(B) shows photograph of RCC porcupine
laid at site. Photo 7.7(C) shows the porcupines laid across a
large channel to induce sedimentation.

Fig 7.7 (A) : Sketch of typical RCC porcupine

Photo 7.7(B) : RCC Porcupine laid at site

Photo 7.7(C) : Porcupine screen laid across a channel

_____

173
 Chapter 8

FILTERS BELOW STONE

PROTECTION WORKS

Percolation of water into the ground or embankment

takes place, causing rise in water table and / or saturation of the
body. The percolation results in seepage force in the form of pore
water pressure. Thee seepage may result in to in piping through
the body of an embankment in the long run. If measures are not
taken to drain the water safely out of the body, failure of bank or
embankment slopes during floods can take place very quickly.
In the preliminary stage, signs of such failure can be seen as
slips, sink holes, boils, settlement, etc.

The desired smooth and efficient drainage of water is

achieved by the use of a layer of more permeable material called
filter.

8.1 CONCEPT OF FILTER DESIGN

The material used for filter should satisfy the following

normal criteria

a. Filter material is more pervious than the base material /

core material

b. Filter material has such gradation that the particles of

base / core material do not migrate through. Such
migration can clog the voids and fail the very purpose of
a filter.

c. Filter material helps to form natural graded material in

the zone of base material / core material.

d. Normally, natural material to satisfy the above require

ments is not readily available in nature. A graded filter
design consists of layers of different grades.

e. To minimise segregation, each grade of the filter

174
 material is normally very uniform in size.

8.2 FILTER BELOW THE REVETMENT

A filter placed between the revetment / protection works

at the surface and the soil bed below the revetment / protection
works. Main function of a filter below revetment is to allow the
ground water to pass through the filter layer but arrest the
movement of soil particles of the bed material.

Hydraulic conditions of the river, characteristics of bed /

bank material, etc are studied before deciding the properties of
filter. Some of the important variables can be listed as, location
of filter i.e. whether on slope or on bed, hydraulic gradient through
the filter and its direction, characteristic properties of sub-soil,
water permeability of sub-soil and soil tightness, suspended
sediment load in the flow and its characteristics, etc.

8.3 DESIGN OF FILTERS

There are two types of filters in use. (a) Graded filter and
(b) geotextile filter

8.3.1 Graded filter

Use of graded filter has been traditionally used below

the revetment and protection works. The material used in the
granular / graded filters is normally durable. The graded filter
material has good surface contact with both the sub-soil and the
top protection works. Under gradually changing conditions,
sometimes, the graded filter can show "self healing" results.

However, the graded filter requires strict quality control

in design and execution. Procurement of suitable material, and
maintaining the desired uniformity of filter material can be a
problem. Therefore, the quality of the final product is not certain.
Graded filters of more thickness show spreading effect causing
damage to the protection works on the top.

Design of graded filter depends upon many factors, which

are site specific. The designer also can have latitude in the design
of graded filter like total filter, permeability of individual layers,

175
etc. Bridge engineers are suggested to use the IS code No 9429
for the design of graded filters. Due to high flexibility available in
the design, each case is normally tackled separately.

Study of literature indicates simple criteria to design

graded filter using a mixture of suitable grain sizes. The relations
indicated in the literature areas below

D50 (filter)
------------------- < 40
D50 (sub-soil)

D15 (filter)
5 < ------------------- < 40
D15 (sub-soil)

D15 (filter)
------------------- < 5
D85 (sub-soil)

Instead of undergoing elaborate designs and construction

of graded filter, mixture of 20 mm to 10 mm stone metal or
brickbats in 75 mm to 100 m thickness is laid below the protection
works. Sometimes kankar in the same range of sizes, screened
from river sand, is used. In many cases, such short cuts have
proved even damaging to the protection works. In such cases
simple relations indicated above can be applied and the filter
material can be made.

8.3.2 Geotextile filter

Geosynthetics is a collective term applied to thin, flexible

sheets incorporated in or about soil to enhance its engineering
performance. The word 'geo' implies the end use associated with
improving properties and performance of engineering works
founded in soil, whereas, the word 'synthetics' implies that the
product is made from man-made polymers. Allied products related
to geosynthetics are geomembranes, geotextiles, geogrids,
geomeshes, geonets, geomats, etc, which are applied in the
foundations under various conditions.

176
 Geotextiles are textile fabrics, made up of polymers,
and are permeable to water. The fine holes or pores in the
geotextile allow the fluids to flow. Due to tensile and warp strength
of the geotextile, geofabric filter has an advantage In case of
uneven settlement of small magnitude. However, in case of
significant settlements, damage to the filter is certain. Any type
of damage to the filter is difficult to repair.

Comparatively the design and construction of geotextile

filters is much simple. The concept of geotextile filters is
comparatively new and not much known to the bridge engineers.
Therefore, The geofabric filters, their designs, etc is discussed
below in details.

8.4 DESIGN CONCEPT

Geofabric textiles are divided in to two types : woven

and non-woven. The woven textiles are made by traditional weaving
methods in which parallel sets of filaments are interlaced
orthogonal to form a coherent textile structure. The size of pores
in a woven fabric is pre-determined. Photo 8.1 shows internal
structure of typical non-woven and woven geofabric filter.

Photo 8.1 : Internal structure of typical

non-woven and woven type geo-fabric

When the soil particle size is very fine, the pore size,
worked out by designs, can be less than that provided by the
weaving system. In such cases, non-woven geotextile is used.
The non-woven textile is made up of fine and long filaments or
fibers of different lengths. The material is laid in specific thickness
and are "needle punched" to mix them thoroughly and form a

177
fully interlaced layer and pressed to form a textile of uniform
thickness. A strong bond between the fibers is formed in the
process.

8.4.1 Properties

A variety of material are used and processes followed in

manufacturing of the geotextile results in a range of physical,
hydraulic and mechanical properties. These are briefly discussed
below

Physical properties normally cover the thickness

(measured in mm), mass per unit area (measured in grams/sq
m) and flexibility, which is measured as length of 25 mm strip to
give a predetermined deflection at its free end.

Mechanical properties normally cover durability, tensile

strength, puncture and burst strength, tear strength and frictional
properties. Durability covers ageing due to oxidation, chemical
attack by acids and alkaline materials found in soil and water,
etc. Tensile strength is uniaxial tear strength of the geotextile
specimen. Tensile strength is normally given in N /cm2 or kN /
m2. Puncture and burst strength measure the resistance of the
geotextile specimen normal to the plane of the specimen. Tear
strength is actually tear propagation strength, measuring the
resistance of the geotextile specimen to tearing after a rupture is
developed in the edge of the specimen. Frictional properties of
the geotextile specimen help to transfer the load from soil to the
geotextile and vice versa.

Hydraulic properties of the geotextile influence its ability

to function as filters or drains. The most important hydraulic
properties are the size and distribution of pores and water
permeability. Just like grain size distribution curves are prepared
for soil, pore distribution curves are prepared for the geotextile
filter. Permeability is the ability of the geotextile to transmit water
across its thickness normal to the plane of the geotextile.

The designer is normally required to decide the

specifications and desired properties, and select the most suitable
product, which satisfy the desired standards.

178
8.4.2 Design criteria

The following simple criteria may be used to select the

correct filter fabric :

(a) Granular material containing 50% or fewer fines (0.074

mm) by weight the following ratio must be satisfied :

% passing size of bed material (mm)

----------------------------------------------------- > 1
opening (pore) size of fabric (mm)

In order to reduce the chances of clogging, no fabric is

specified with an equivalent opening size smaller than 0.149 mm
and equal to or less than 85% passing size of the bed material.

(b) For bed material containing at least 50% but not more
than 85% fines by weight, the equivalent opening size of filter
should not be smaller than sieve No.100 (0.149 mm) and should
not be larger than sieve no.70 (0.211 mm).

(c) For bed material containing 85% or more fines, filter

should not be provided directly on the bed material. A cushion of
fine sand of thickness 10 cm is necessary below the geofabric
filter.

It may be added here the technology of producing woven

fabric of finer mesh is developing fast. Therefore, the simple
criteria indicated above can be superseded in a short period.

8.5 METHOD OF CONSTRUCTION

10 cm thick layer of sand having D50 between 0.2 to 0.6

mm is to be laid between base layer and filter. The size of filter
opening should be smaller than D85 of sand layer. It would be
preferable to use a fabric with an opening as large as allowed by
the criteria mentioned above.

Over the synthetic filter, a 15 cm thick layer of coarse

sand and / or gravel should be provided before placing of stones.
Photo 8.2 (A), 8.2 (B) and 8.2 (C) shows the geofabric filter being
laid below stone pitching at site.

179
Photo 8.2 (A) : Showing geo-fabric being laid below the
slope pitching

Photo 8.2 (B) Geofabric filter being laid below the stone
pitching. Cushion of sand is laid before spreading the
geofabric filter.

Photo 8.2 (C) : Geofabric filter being laid below the

stone pitching. Cushion of sand is laid on the top of filter.

180
REFERENCES

8.1 Geotextiles and Geomembranes Manual, Dr T S Ingold,

Elesevier science Publishers Ltd Mayfield House, Oxford,
OX2 7DH, U K

8.2 Geotextiles and Geomembranes in Civil Engineering,

Edited by Veldhuijzen Van Zanten, A. A. Balkema /
Rotterdam / Boston / 1986

8.3 Drainage system for earth and rockfill dams", IS

document No 9429 :1999 - First revision

181
 Chapter 9

HYDRAULIC MODELLING

9.1 MODELLING TECHNIQUE

Modelling is neither new nor relevant to hydraulic

engineering only. It is practiced in the field of hydraulic
machinery for testing pumps and turbines, in ship
hydrodynamics for studying ship characteristics and in
aircraft industry for testing aerofoil shapes. Modelling can be
either physical or conceptual. Hydraulic and structural
models are illustrations of physical modelling while
mathematical and stochastic models are examples of the
second category. Hydraulic modelling has become popular
because of its several inherent advantages. Design conditions
are rarely obtained in nature. Even if they are sometimes
experienced, their duration can only be short. It is, however,
possible to impose design conditions on the model any
number of times and sufficiently long to permit thorough
testing. Models are also useful in diagnosing causes of
failures. New ideas can be easily put to test in a model but
not in a proto-type. In view of these over whelming benefits,
hydraulic modeIIling has been accepted and practiced
universally.

9.2 SIMILITUDE REQUIREMENT

When model is required to provide an aid merely for

visualisation, scaler reproduction of form is sufficient which is
termed Geometric Similarity. If in addition reproduction of
fluid motion is desired, laws of Kinematic Similarity are
required to be followed. Study of hydrodynamic fluid forces
is, however, more useful when Dynamic Similarity is required
to be achieved in the model, conditions for which are two
fold. One is that the ratio of each one of the forces in the
model to that in the prototype should be the same. The other
condition is that in both the model and the prototype, sum
total of all forces should be equal to inertia force.

182
 With turbulent flow in channels, gravity force
predominates. Ignoring other forces and applying the above
conditions, it is then found that the Froude Number (Fr) equal
to V/(gL)1/2 should be the same in model and prototype
wherein V is the mean velocity, g is the gravitational
acceleration and the L the length parameter. In laminar flow
phenomenon, the viscous flow predominates and the
conditions obtained is that Reynolds Number R equal to VL/
v wherein v is the kinematic viscosity should be the same.
Surface tension achieves importance in certain situations
such as weir flow with smalI depths. The condition that then
emerges is equality of Weber Number (W) given by
V2 (σ l δL). In this σ is the surface tension and δ the density.
In water hammer problems elastic compression is an
important force. Couchy Number C equal to V/(K / δ)l/2 should
then be same in model and prototype in which K is the bulk
modulus of elasticity. Each of the variables in above
quantities should have same basic fundamental dimensions
of length, time and mass / force so that the number (Froude,
Reynolds, etc.) is dimensionless.

9.3 MODEL SCALE DESIGN

Knowledge of dynamic similarity conditions enables

evaluation of model scales. Almost alI the problems related
to river hydraulics are about gravity flow and hence required
to follow the condition of equality of Froude Number. In
evolving model scales for such problems it is logical to
classify models of different types and consider them
separately. First broad classification is according to nature of
boundary- rigid boundary and mobile boundary models. Next
grouping in each category can be according to geometry -
whether geometricalIy similar or dissimilar on account of
vertical exaggeration.

9.3.1 Rigid bed models

Models with rigid boundary and geometrically similar

form are used in studies connected with outlet works, stilling
basins, spillways etc. Conditions of Froude Number equality,
equality of horizontal and vertical scales and adoption of
some flow formula enable working out of the scales of the

183
model. The scale of boundary friction in this case works out
to be such that model is required to be made smoother than
the prototype.

Rigid boundary models are often constructed for

determination of flood levels in rivers and flood routing
studies. If geometrically similar model is constructed, depth,
velocity and Reynolds Number in the model becomes too
small whereas discharge required becomes too big. These
difficulties can be surmounted if depths are exaggerated in
the model. Such models are, therefore, termed vertically
exaggerated models. Since long lengths of rivers are
normally required to be reproduced in such models,
availability of space usually puts constraint on horizontal
scaling. Vertical exaggeration results in model roughness
being required more than in the prototype which is often
achieved by embedding artificial roughness elements. This
puts another constraint on scaling. Exaggeration of depths in
relation to horizontal dimensions is generally high in such
models.

9.3.2 Mobile bed models

Mobile bed models can again be either geometrically

similar or vertically exaggerated. Models of the former type
permit study of problems of local scour at piers, abutments,
spurs, etc. Model scale design is similar to that in case of
rigid boundary model except the equality of T/Tc ratio is an
additional condition to be satisfied for reproduction of
sediment transport. This simple approach is based on the
concept that sediment transport is governed primarily by T
and Tc values.

In river models where bed changes are required to be

studied, mobile bed vertically exaggerated models are used.
Scale design in this type of models can be made on basis of
Lacey formulae for width, depth and slope. Alternatively more
rational procedures due to Einstein(9.2), Grade(9.3) and others
endeavor to satisfy several conditions in addition to Froudian
and flow resistance laws for reproduction of specific aspect of
sediment transport phenomenon such as incipient motion,
bed forms, bed load transport suspended load distribution,

184
local scour, channel pattern(9.4) etc. Vertical exaggeration in
river models normally found suitable on basis of above
conditions is about 3 to 5.

9.4 MODEL LIMITATIONS

Due to vertical exaggeration, model slope

automatically becomes steeper than in the prototype, this
slope exaggeration being the same as the vertical
exaggeration. When river has a very flat slope even such
exaggerated slope in the model may be found to be
insufficient to cause satisfactory bed load movement. The
model is then made steeper than the slope obtained due to
vertical exaggeration and it is called tilted. Complications
arising out of this expendient like non-reproduction of
backwater heights and lengths are required to be taken care
of separately. Complications arise even due to vertical
exaggeration alone. Scour holes are not correctly formed
since side slope of scour hole in the model cannot become
steeper as angle of repose of bed material can’t be very
different. Weir coefficients become higher in the model since
shape of weir crest is vertically exaggerated. Pier contraction
effect becomes excessive in the model because the height of
pier is exaggerated. Suitable corrective measures are
introduced to overcome such difficulties and where this is not
possible, allowance for divergence in the model from the
proto type is made in interpreting model results.

9.5 DIMENSIONAL ANALYSIS

Modelling technique is sometimes resorted to for

finding out the general law of physical phenomenon such as
flow over a weir or scour at piers. Dimensional analysis
provides a simple and rational approach in such studies with
the use of the p theorem. In general a system may involve n
variables which can be expressed in terms of m dimensional
units. Buckingham’s p Theorem states that the general
physical law can then be expressed as a function of n-m
dimensionless p terms, and each of such p terms will have
m+1 variables of which only one may be changed from term
to term. Once the relevant p terms are determined and the
structure of functional relationship governing the physical

185
phenomenon is conceived, model experiments can be
programmed and conducted to determine the coefficients and
indices entering into this relationship.

Besides scaling of models, several other features

such as model construction, model operation,
instrumentation and interpretation of model results are also
important and all of them need careful attention.

Important literature on the subject is cited under

references.

9.6 REQUIREMENT OF FIELD DATA

Important design features of major bridge projects

should be tested and finalised by hydraulic model studies.
Bridge waterway, shape, length and layout of guide bunds,
along the bridge and approach banks, distribution of
discharge in various spans of the bridge, additional training
works required if any are such aspects which are normally
studied in a hydraulic model. Field data is required to be
collected for this purpose which is listed below(9.10).

9.6.1 General

Preliminary report of the bridge project.

9.6.2 Survey data

(i) Plan of the river extending over 2 meanders

upstream and one meander downstream of bridge site. In
case of a straight reach length of river reach may be at least
4 times the khadir width on upstream and 2 times on
downstream. Tentative bridge location, guide bund layout,
approach banks, marginal embankments, existing training
works if any etc. should be shown on this plan. Edges of low
water channels should be marked.

(ii) Cross sections on bridge line, at upstream, down

stream· boundaries and in the intermediate reaches. Spacing
of sections should be such that in the model, distance in
between them would not be more than about 3 m.

186
 Positions of sections and their zero chainages
should be shown on the plan. The levels should be close
enough to enable demarcating deep and shallow portions of
the channels but can be wide apart where variations in levels
are gradual.

Cross section at bridge alignment should be

observed at low water, medium flood and at near high flood
stage.

(iii) Longitudinal section of the major channel along deep

portion, termed Thalweg, Position of L section line should be
shown on the plan giving periodical changes.

9.6.3 Hydrographic data

(i) Design discharge for bridge waterway and foundation

with corresponding high flood levels.

Frequency curve for annual peak discharges.

(ii) Gauge-discharge or rating curve at bridge site

showing rising and falling limbs separately.

(iii) Hydrographs pertaining to normal year and also the

one containing peak discharge for foundation design.

(iv) Rating curves should also be observed near the

cross section at model boundaries. Alternatively low
water and high water slopes should be reported.

9.6.4 Sediment data

(i) Details about weighted mean diameter and silt

factors or river bed material on bridge section at the
positions of abutments and piers. This data should
be observed at river bed and at various depths below,
down to maximum estimated scour depth.

(ii) Engineering properties like cohesion, angle of friction,

Atterberg limits etc., of bank material under
saturation conditions during falling floods. This data

187
 should be observed near bed level, at mid depth and
near H.F.L. on bridge line.

(iii) Suspended sediment size and concentration at

medium and high flood stages.

REFERENCES

9.1 American Society of Civil Engineers ‘Hydraulic

Models’, Manual No.25, 1942.

9.2 Einstein, H.A. and Nihg Chien, ‘Similarity of distorted

river models with movable beds’, Transactions,
American Society of Civil Engineers, Volume 121,
1956, pp 444-462.

9.3 GARDE, R.J. ‘Analysis of distorted river models with

movable beds’, Irrigation and Power Journal, Central
Board of Irrigation and Power India, October” 1968,
PP 421"-432.

9.4 Garde, R.J. ‘Physical modelling of alluvial river

problems’, proceedings of the international Workshop
on Alluvial River Problems, Roorkee, U.P.lndia,
March 1980, pp ‘1-1 to 1-14. 6.5 Hunter Rouse,
‘Engineering Hydraulics’, Wiley and Sons, Chapter II,
pp 136-228.

9.6 United States Bureau of Reclamation, “Hydraulic

Laboratory Practice,’Engineering Monograph, March
1953, Revised 1981.

9.7 Allen, J., ‘Scale models in hydraulic engineering’,

Longman’s Green & Co., London, 1947.

9.8 Dandekar, M.M., ‘Hydraulic Similitude’, Journal of

Institution of Engineers (India), Nov. 1963.

9.9 Central Water and Power Research Station, Pune,’

Methods followed in fixing model scale ratios and
overcoming model limitations based on experience
gained at the Indian Water ways Experiment Station,

188
 Pune’. Annual Report (Technical), 1945, pp 88-116.

9.10 Central Board of Irrigation and Power, India, ‘Draft

Manual on type of field data required for model
experiments’, August 1964.

9.11 Central Board of Irrigation and Power, India ,’Manual

on river behaviour, control and training.,’ Publication
No,60, Revised 1971, Chapter X. pp 354 –396.

189
 Chapter 10

MEASUREMENT OF RIVER WATER

LEVELS AND DISCHARGES
In designing bridges and other engineering structures
on rivers, maximum discharge and highest flood level form
essential data for design purposes. If the structure is meant for
harnessing river water supplies, discharge and water level
observations may be required to be made more often and in certain
cases even daily. Various methods used in observation of water
levels and measurement of discharge are described below, based
primarily on the Indian Standards cited under reference. However,
for making “short term” and “long term” measurement of discharge
in accordance with Khosla Committee’s recommendations,
proposed Manual by R.D.S.O. for “Estimation of Design Discharge”
may be referred. For methods of gauging, measuring rainfall and
taking discharges on bridges nominated by R.D.S.O., instructions
issued by R. D. S. O. may be followed.

10.1 WATER LEVEL OBSERVATIONS

10.1.1 Site requirements

Site for installing gauges should be properly selected.

The reach should be reasonably straight and of uniform cross·
section and slope. The length of such a reach should not be less
than 400 m and need not be more than 1600 m. When the length
is restricted upstream of the measuring section, it should be
double of that on the downstream side. When near a confluence,
the site, if on a distributory, should be beyond backwater effect
and if on the mainstream, should be beyond the disturbance due
to tributary or structures like weirs. The site should be easily
accessible even during floods.

10.1.2 Different types of gauges

The gauges can be either of the two categories, non-

recording or recording types. The simplest non-recording gauge
is the staff gauge. It can be erected vertical in one piece or often
in pieces of one or two metres each suitably spaced so that

190
there is some overlap of the successive gauges. In order to obtain
more accuracy, gauges can also be fixed sloping if site conditions
permit.

Chain, wire and tape gauges are fixed on structures like

bridges and the weight is lowered to touch the water surface.
Movable stilling well and electrical arrangement for detecting
contact of weight with water are useful gadgets in this type of a
gauge.

Pneumatic gauges permit reading or recording water level

at some distance from the river, say up to about 305 m. In a
diaphragm type gauge, water presses against metalIic diaphragm
fixed to a metal air bell. Air inside the bell changes its pressure
with the variation in depth of water column above the diaphragm
the air belI is connected by tubing to a dial indicator which can
be located in a cabin at a safe distance from the bank. In a
bubble type pneumatic gauge, the bell is provided with a slot.
The air pressure is read by manometer and converted to depth of
water and thus the water level is obtained. Electrical long distance
gauges are also available of different forms.

In recording type gauges, automatic float operated

gauges are more common. The well for the float provides room
for installation of the recorder, working of float and a counter
weight. Outside disturbances are damped by restricting the cross
sectional area of the inlet pipe to 1/1000 of the inside horizontal
area of the well. To minimize errors arising on account of lag of
stylus and due to submergence of float, care is required to be
taken to reduce friction of using large size float and light weight
float line. More details of the float well are available in reference
number 10.26.

Pneumatic gauge installation does not need a gauge

well. Bubble gauge of this type gives precise observations, covers
great range of variation in water level and operates over long
periods up to three or more months continuously without
attendance. Electrical long distance gauge recorders and radio
telemetering system are also available.

191
10.1.3 Crest gauges

In case of bridges, it is sometimes necessary to observe

high flood levels rather than daily gauges to ensure that minimum
clearance is not encroached upon. Very high peaks may occur
any time during day and night. Unless recorder type gauges are
installed it is difficult to obtain peak gauge data. Devices called
crest gauges are useful to record peak flood levels under such
conditions. These can be of several types as indicated in Fig.
10.1 and 10.1 A

A simple crest gauge consists of bottles fixed to a staff

gauge. Another version is an empty pipe with inside staff gauge
on which peak water level is marked by granulated cork powder
deposited inside the pipe. One more type of crest gauge has a
float moving inside a transparent tube, the float being prevented
from moving down with receding water level by means of a rubber
skirt attached to it. Another simple device is to paint stripes on
the staff gauge by yellow dextrine powder which leaves a high
water mark when it gets dissolved in water. Details of such peak
water levels over a period of years collected at the -bridge site
will help in flood frequency studies.

10.2 DISCHARGE MEASUREMENTS

Methods of discharge measurement can be broadly

divided into four categories: velocity-area method, slope-area
method, stage discharge relationships and other methods.

10.2.1 Velocity - Area Method

Out of the above four categories, velocity area method(10.1)

is most widely used and provides more accurate results.
Requirements of a good site in this method are more or less
similar to those in case of gauge observations. The discharge
site after selection is marked by means of masonry or concrete
pillars on both the banks.

Procedure adopted in this method is of observing depths

and velocities on a number of verticals across the section.
Measurement of depth is made at intervals close enough to define
the cross sectional profile accurately. Velocity observations are

192
Fig. 10.1 : Staff - in - pipe type crest - stage gauge

193
Fig. 10.1 A : Crest gauge

194
normally made at the same time and on the same verticals used
for depth measurements. It is considered safe to normally use
25 equidistant verticals for depth and velocity observations. In
the case of a river, the following values of relative errors at 68%
confidence limit are given for progressively reduced number of
verticals.

Number of verticals Relative error

25 1.7%
20 2.2%
15 3.0%
less than 10 more than 4.5%

For artificial channels, smaller number of depth and

velocity verticals than in case of rivers can be adopted to restrict
the relative error to 2 percent.

Where the depth and velocity measurements are out of

necessity made at different times, the velocity observations are
made ordinarily on 15 to 20 verticals. The width of segments is
not allowed to be less than 0.25 m and not greater than that in
which the mean velocities in two adjacent verticals differ by more
than 20 percent with respect to the lower value of the two.

Various methods are available for marking segments,

of which pivot point layout is more common which is shown in
Fig. 10.2.

When velocities are lower and depths smaller up to

about 0.9 m, wading observations can be made using wading
rods. When depths vary between 0.9 and 4.6 m, sounding rods
can be used while for still bigger depths hand line consisting of a
cable is convenient to use. When weights in excess of 13.6 kg
are required to be used in deep water, a cable line is raised or
lowered by means of a crane. For depth up to 9.0 m and velocities
up to 3.0 m/s, rack and pinion equipment can also be used in
boat measurements.

Even though sounding weights (10.14) are of stream lined

shape, the drag on them and on the sounding cable causes

195
196
Fig. 10.2 : Pivot point layout
deflection on the suspension line from true vertical in bigger depths
and higher velocities. The conventional air and wet line corrections
are then applied.

Introduction of these errors due to drag and deflection

of cable will become manifest from Fig. 10.3.

The percentage correction which should be deducted

from the measured length of the sounding line above water
surface, called the air line correction, is given in Table 10.1.

The wet line correction, also expressed as a percentage

to be deducted from the measured length of the sounding line
below water surface is given in Table 10.1 for angles up to 30°.

Table 10.1 : Air Line Correction and

Wet Line Correction

Vertical Air Line Wet Line

Angle Correction Correction
4o 0.24 0.06
6o
0.55 0.16
8o
0.98 0.32
10 o
1.54 0.50
12 o
2.23 0.72
14 o
3.06 0.98
16 o
4.03 1.28
18 o
5.15 1.64
20 o
6.42 2.04
22 o
7.85 2.48
24o 9.46 2.96
26o 11.26 3.50
28 o
13.26 4.08
30 o
15.45 4.72

197
 A

L1
L1
SOUNDING LINE

C
K L = AIR LINE CORRECTION
B a
D

L2 CURRENT L2

KL
W WET LINE CORRECTION
RIVER BED E
G

K La = (Seca - 1) L1

Fig. 10.3 : Air line and wet line corrections

198
 o
Significant errors may result if the vertical angle is more
than 30 .

Echo sounders do not need such corrections and

facilitate making depth measurements more easily and quickly.

The echo sounder works on the principle of sending out

a pressure wave, receiving back the echo i.e. reflected wave and
measuring and translating the interval of time in between in terms
of depth. Velocity of pressure pulse is affected due to change in
density. The transducers are, therefore, mounted at a location
and elevation where intensity of stream of air bubbles generated
due to turbulence, wind velocity, launch speed or any other cause
is minimum. The operating frequency of the transducer is chosen
to obviate effect of water noises, minimise loss of energy, minimise
error on account of wide sound cone and to provide measurable
time lag between incident and echoed waves. When echo sounder
is used for measuring depth of scour at pier or along guide bund,
wide cone introduces error. To minimise this error, sonic
frequencies are avoided and supersonic range giving narrow sound
cone is preferred. The echo sounder can be either indicator or
recording type. Accuracy obtainable is about ± 2 cm at 15m
depth for the former and ±1 cm in case of the later. Accuracy· is
reduced with lower depths and is affected by change of
temperature. The echo sounder is, therefore, ideal equipment for
measuring big depths especially when velocities are high which
may introduce significant errors in sounding weight measurements
due to deflection of sounding line.

Velocities are measured by current meters either of the

vertical axis cup type or the horizontal axis screw or propeller
type. Of these, the cup type current meter is in common use in
India and the design is very similar to Gurley, Price or Watts
Current meter. IS 3910-1966 gives specifications for cup-type
current meter(10.11) and IS 3918-1966(10.9) gives the code of practice
for use of this type of current meter. Normally the current meter
is exposed at the point of measurement for a minimum of 40
seconds and the velocity is deduced from an average of 3
readings. Rating of the current meter is made in a toeing tank
and repeated after 300 working hours or after a total period of 6
months of use whichever is earlier. Before the meter is used,
spin test is made and the minimum spin of 90 seconds without

199
and 75 seconds with spring contact is ensured. Under ideal
conditions the cup type meter is expected to give accuracy up
to 98 percent.

Several methods are in vogue for determining the mean

velocity over a vertical such as one point, two point and multipoint
methods.

In one point method, velocity is observed at 0.6 depth

while in two point method, two observations are made; one at
0.2 depth, second at 0.8 depth and their average taken as mean
velocity. Both these methods are in common use though one
point method is mostly used in India. Subsurface method entails
making velocity observation just below the surface velocity by
reduction coefficient which varies from 0.79 to 0.91 and depends
on depth, slope and relative roughness of bed material. This
reduction coefficient is determined for a particular site by actually
observing surface and mean velocities during low and medium
flood stages by field observations and drawing a curve of stage
against reduction coefficient and extrapolating it for desired high
flood stage.

For computing discharge, mid section method is preferred

as it gives more accurate results. In this method, the discharge
intensity on the observation vertical is deemed to be operative
over half the adjacent segments. With this assumption, discharge
is computed in different parts of the cross section and added up
to obtain the total discharge. Correction is applied in working out
the discharge if the water level varies during the period of discharge
measurement.

Ordinarily boat observations are made for measuring river

discharge. Cableway can also be erected for stream-gauging
purposes, especially in deep river gorges where current is fast
and depths too big. Bridge observations are preferred when the
river has considerable spills and facility for observations using a
survey launch not available or the river is otherwise not accessible
during floods. The velocity in this case is measured from
downstream parapet of the bridge and more number of verticals
are adopted in view of variation in discharge distribution across
the section on account of pier obstruction.

200
 While measuring discharge from the bridge, according
to Research, Designs and Standards Organisation, the number
of verticals should be as indicated below:

Length of Span Number of verticals

in each span
Up to 6 m 1
6 to 12 m 2
more than 12 m 3

The number of verticals are suitably modified depending

on time required for completing one set of observations and
duration of flood assessed by enquiry or past experience.

The effect of obstruction due to piers can be allowed for

by locating the verticals adjacent to the piers judiciously in the
following way, depending on whether the river bed is nonscourable
or scourable.

Owing to obstruction, the velocity field gets affected in

the vicinity of the piers. The flow on either side of the pier becomes
accelerative. When bed is inerodible scour does not however,
form round the pier. At the time of making velocity observations,
it is, therefore, necessary merely to avoid this region of
accelerative flow and locate the positions of verticals adjacent to
the pier sufficiently away. The discharge intensity 'q1' on the vertical
'L1' located at this distance may be assumed to be operative over
the region 'W1' between the verticals and the centerline of the
adjacent pier. The discharge in the width between vertical 'L1' and
the face of the pier will, therefore, be equal to 'q1 W1'.

In case of rivers permitting scour hole to develop round

the piers, the flow field covering scour hole gets affected. The
discharge into and out of the scour hole is required to be adjudged
and allowed for. This can approximately be done by assuming
the velocity and discharge intensity 'q2' on the vertical 'L2' beyond
the scour hole on the gauging station to be operative in the width
'W2' between the vertical 'L2', and face of the pier. The discharge
in the width between vertical L2 and the face of the pier will then
be obtained as 'q2W2'. The extent of scour hole in plan can be

201
estimated on basis of depth of scour assuming angle of repose
of sand as 2 horizontal to 1 vertical.

Standard forms are used for record of gauges, cross

section, computation of discharge etc. These are published in
ISI 1194-60(10.3). Alternatively the forms devised by the PCC set
up pursuant to recommendations of Khosla Committee may be
used.

10.2.2 Slope-Area Method

In the event of infeasibility of Velocity Area method on

account of either rapid rise or fall of floods or lack of equipment
or any other reason, the Slope-Area method is adopted for rough
estimation of discharge.

The requirements of site are mostly similar to those in

case of Velocity-Area method. In addition, the length of the reach
in Slope-Area method is required to be reasonably stable, free
from obstruction, straight and of uniform section for a length not
less than 5 times the width of the channel and in any case not
less than 300 m. The surface drop in this reach is also required
to be not less than 150 mm. The gauges are read to 2 mm and
the slope is worked out from the average of gauge observations
at either end of the reach.

The cross sectional area is measured adopting the

procedure as in case of Velocity-Area method. The velocity
formula commonly used is that of Manning (Chapter 4.1), the
slope entering the formula being the energy slope which accounts
also for the kinetic head difference. The rugosity coefficient value
is adopted following Tables 10.2 and 10.3 for coarse bed material
and for bed material other than coarse respectively recommended
by the Indian Standards Institution.

202
 Table 10.2 : Values of Rugosity Coefficient n for open
channels with Relatively Coarse Bed Material
not characterized by Bed Form

Sr. Type of Bed Size of Material Rugosity

No. Material Equivalent Coefficient n
Diameter
1 Gravel 4 mm to 8 mm 0.019 to 0.020
8 mm to 20 mm 0.020 to 0.022
20 mm to 60 mm 0.022 to 0.027

2 Cobbles 60 mm to 110 mm 0.027 to 0.030

and Shingle 110 mm to 250 mm 0.030 to 0.035

Table 10.3 : Values of Rugosity Coefficient n for open

channels with other than Coarse Bed Material

Types of Channel and Description Rugosity Coefficient ‘n’

Excavated or Dredged Channel
(a) Earth, straight and uniform:
(i) Clean recently completed 0.016 to 0.020
(ii) Clean after weathering 0.018 to 0.025
(iii) With short grass, for weeds 0.022 to 0.033
(b) Rock cuts :
(i) Smooth and uniform 0.025 to 0.040
(ii) Jagged and irregular 0.035 to 0.050
Natural streams
Minor streams (Top width at
flood stage less than 30 m),
(i) Streams on plains-clean, 0.025 to 0.033
straight, full stage no rifts
or deep pool
Flood on plains :
(1) Pasture no brush
(i) Short grass 0.025 to 0.035
(ii) High grass 0.030

203
 (2) Cultivated area
(i) No crop 0.020 to 0.040
(ii) Mature raw crops 0.025 to 0.045
(iii) Mature field crops 0.030 to 0.050
(3) Brush
(i) Scattered brush, 0.035 to 0.070
heavy weeds
(ii) Light brush and trees 0.035 to 0.060
(without foliage)
(iii) Light brush and trees 0.040 to 0.080
(with foliage)
(iv) Medium to dense brush 0.045 to 0.110
(without foliage)
(b) Medium to dense brush 0.070 to 0.160
(with foliage)
(4) Trees
(i) Cleared land with 0.030 to 0.050
tree stumps, no sprouts
(ii) Same as above but with 0.050 to 0.080
heavy growth of sprouts
(iii) Heavy stand of timber, 0.080 to 0.120
a few down trees, little
under growth flood stage
below branches
(iv) Same as above but with 0.100 to 0.160
flood stage reaching
branches
(v) Dense willows, straight 0.100 to 0.200

IS 2912-1961 (10.4) give more detailed recommendation

for adoption of the Slope-Area method.

10.2.3 Stage Discharge Relationship

Regular recording of discharges over a period of time is

essential for estimation of water resources of river basins and
subsequent planning and utilisation. Daily discharge observations
over a long period are sometimes not feasible being too expensive.
The estimation of discharge is then achieved by using proper
stage discharge relation.

204
 Besides the site requirements mentioned in connection
with Velocity Area method, the following additional requirements
are required to be satisfied in adopting the method of Stage
Discharge relationship.

The site should provide a natural physical control for

obtaining a stable stage-discharge curve. In the absence of such
a control the site should provide at least partial control over the
major range of stage discharge curve. If natural control is not
available and the stream is sufficiently small, an artificial control
in the form of a low sill or notch can be provided.

As far as possible the site should be free from variable

back water effect. If such a site is not available, necessary
correction is applied using twin guage method mentioned later.
The site is so selected as to give a significant change in stage
with change in discharge. This requirement ensures sensitivity
of the stagedischarge curve which is also termed as rating curve.

The rating curve is constructed on basis of observed

gauge and discharge data by fitting a mathematical curve. A
sufficient number of discharge measurements using Velocity-
Area method are made over the range of variation of gauges. The
number of observations are so made as to evenly distribute the
plotting points throughout the whole range of the rating curve.
The observations made at rising and falling stages are indicated
separately. In erodible channels, the discharge observations are
made at frequent regular intervals and the rating curve is checked
and adjusted from time to time.

The rating-curves are tested for absence from bias, for

goodness of fit and for shift in control. In a bias free curve an
equal number of observations are expected to be above and below
the curve. As a rule this is not allowed to be greater than the
percentage standard deviation of the error of the discharge
observations.

Usually, the gauge discharge relationship is expressed

by the equation Q = C (G-Go)n wherein Q is the discharge, C is a
constant, G is the gauge height, Go is the gauge height for zero
discharge and n is another constant.

205
The value of Go is determined by using the relation

G1G3 − G22
GO =
G1 + G3 − 2G 2
Wherein G1, G2 and G3 are the three gauge heights
corresponding to three discharges Q1, Q2 and Q3 which are
selected in geometric progression. Alternatively the standard
graphical method can be used.

If control does not change, the extrapolation of the stage

discharge curve over a limited range is considered permissible.
Separate extrapolations of stage-area and stage-velocity curves
are also made to obtain extrapolated value of the discharge. After
the bankful stages the discharge of the spill portion is worked
out by estimating the velocity separately for the spills. Another
method used for extrapolation when second discharge site exists
on the same stream is to assess the relation between the rating
curves of the two stations.

Establishment of a single gauge station is not sufficient

where the flow at the station is affected by conditions existing
either upstream or downstream.

In such cases, twin gauge method is adopted by

establishing two gauge stations, distance between them being
sufficient to ensure adequate difference in water levels, not less
than 0.1 m. The normal fall and constant fall methods are then
used for obtaining the discharge in case of variable slope due to
backwater effect. In these methods, additional third variable of
fall between the twin gauges is used. In normal fall method, this
fall is the one which separates the region of noticeable effect of
back water in the plot of gauge-discharge data from that having
no effect of backwater. In constant fall method, any suitable fixed
value of fall is adopted as a third variable.

Correction is required to be applied when flow is unsteady.

The true discharge of an unsteady flow is worked out from the
normal steady discharge Qo obtained from stage-discharge curve
using the following formula

206
 1
⎛ 1 dh ⎞ 2
Q = Q O ⎜1 + . ⎟
⎝ SO VW dt ⎠

wherein So Is the normal slope corresponding to steady discharge,

Vw the wave velocity and dh/dt the rate of change of stage with
respect to time.

Forms for use of record of gauges, for computation of

discharge using Velocity-Area method, daily statement of
"discharge observations, daily stage discharge data, etc. are
standardised.

Exhaustive instructions for adopting the method of

discharge estimation by establishing stage-discharge relationship
are contained in IS 2914-1964(10.6).

10.2.4 Other methods

Moving boat method has been more recently evolved

and tested in the U.S.A. and holds promise in measurement of
discharge of big rivers and tidal waterways. If a current meter is
towed by means of a boat in the same direction as the current,
the velocity measured by the current meter will be the sum of
boat velocity and-the current velocity. If the current meter is towed
in the direction opposite the current, velocity measured by the
current meter will be the difference between the boat and the
current velocities. By moving boat parallel to flow at various
distance from the bank, velocity of flow across the section can
thus be computed. This is the underlying principle of moving
boat method. Since the whole operation takes relatively short
time, the method is convenient for adoption in case of big rivers
and tidal waterways. Further details of the methods are available
in literature.(10.26)

Dilution methods are more suited especially in hilly

streams and torrents where current meter measurements are
not feasible on account of excessive velocity and turbulence,
small depth of flow, moving gravel and coarse materiel etc. In
chemical dilution method, the chemical may be injected suddenly
only once or it may be injected continuously. The former is called

207
sudden injection method and the latter constant injection method.
The apparatus and chemical to be used, specifications of
techniques of injection, sampling and also the methods of
analysis are covered in reference number 10.25.

Radio tracer technique is similar to chemical dilution

methods. A known quantity of suitable radio tracer is dumped
and after it is thoroughly mixed some distance, downstream, the
number of click caused by the tracer in its passage are monitored
on the Geiger counter at the downstream station on the river
bank. If the total number of counts is N, it varies inversely as the
rate of flow Q since flow with lower velocity will allow more time
for counts to accumulate on the register. On the other hand N
varies directly as A, the quantity of radio tracer. Thus N = AF/Q
or Q = AF/N. In this expression F is the proportionality factor
which is the characteristic of the isotope, the counter and the
geometric relationship of the stream. Value of F can be determined
by laboratory tests(10.26).

In acoustic or ultrasonic method, transducers are fixed

on opposite banks of the river, position of one of the transducers
being upstream on the river. They generate acoustic pressure
pulse which travel across the section. The difference between
speeds of ultrasonic waves in and against the direction of water
is measured from which the· flow velocity can be computed on
basis of which discharge is worked out.

In electromagnetic method, a magnetic field is generated

by the electric current flowing through a coil fixed below or above
the channel. The voltage induced in the flowing water by the
magnetic field is related to discharge. Once this relationship is
determined, observation of voltage permits direct estimation of
discharge.

Existing structures like weirs(10.2), (10.23), sluices(10·8), drop

structures(10.21), (10.22) can be conveniently used for measurement
of discharge. Formulae for flow over or through such structures
are well known. Considerable work has been done to determine
variation in the coefficients entering the formulae due to differences
in the geometry and hydraulic conditions obtained at individual
structures. Some times structures like standing wave flumes
(10.18), (10.19)
sharp and broad crested weirs (10.20) etc. are constructed

208
specially for metering the water supplies. Rigid specifications
about their shapes and approach and exit conditions, locations
and method of measuring water levels, formulae and coefficients
to be used are then required to be followed. Reference numbers,
10.2, 10.8, 10.18, 10.19, 10.20 and 10.21 deal exhaustively with
standard notches, weirs, flumes, falls, drops, orifices, nozzles,
ventury etc.

REFERENCES

10.1 IS: 1192-1981, Indian Standard- Velocity-Area methods

for measurement of flow of water in open channels (First
revision)
10.2 IS: 1193-1959, Indian Standard-Flow of Water in open
channels using notches, weirs and flumes.
10.3 IS: 1194-1960, Indian Standard-Forms for recording
measurement of flow .of water in open channels, June
1960.
10.4 IS: 2912-1964, Indian Standard - Recommendations for
liquid flow measurement in open channels by Slope-
Area method (approximate method) (Amendments 2)
10.5 IS: 2913-1964, Indian Standard - Recommendation for
determination of flow in tidal channels (Amendment 1).
10.6 IS: 2914-1964, Indian Standard - Recommendations for
estimation of discharge by establishing Stage- Discharge
relation in Open channels (Amendment 1)
10.7 IS: 2915-1964. Indian Standard - Instructions for collec
tion of data for the determination of error in measure
ment of flow by Velocity-Area methods (Amendment 1)
10.8 IS 2952 (Part I) - 1964, Indian Standard - Measurement
of fluid flow by means of Orifice plates and nozzles,
Part I - Incompressible fluids.
10.9 IS: 3918-1966, Indian Standard - Code of Practice for
use of current meter (cup type) for water flow measure
ment (Amendment 1)
10.10 IS : 4477· (Part I) - 1967, Indian Standard-Methods of
Measurement of fluid flow by means of ventury meters,
Part I, Liquids, May 1968.

209
10.11 IS: 3910 - 1992, Indian Standard - Requirements for water
flow current meters. Measurements in open channels
rotati'lJ elements.
10.12 IS: 3911 - 1994, Indian Standard - Surface, floats,
functional requirements.
10.13 IS : 3912 - 1993, Indian Standard - Sounding rods
functional requirements.
10.14 IS: 4073 - 1967, Indian Standard - Specification for fish
weights, June 1967.
10.15. IS: 4080 - 1994, Indian Standard - Specification for vertical
staff gauges - functional requirements.
10.16 IS: 4858 - 1968, Indian Standard - Specification for velocity
rods.
10.17 IS : 6064 - 1971, Indian Standard - Specification for
sounding and suspension equipment, December 1971.
10.18 IS: 6063 - 1971, Indian Standard - Method of
measurement of flow ·of water in open channels using
standing wave flume, November 1971.
10.19 IS: 6062 - 1971, Indian Standard - Method of
measurement of flow of water in open channels using
standing wave flume fall, November 1971.
10.20 IS: 6059 - 1971, Indian Standard - Recommendations
for liquid flow measurement in open channels by weirs
and flumes - weirs of finite crest width for free discharge,
September 1971.
10.21 IS: 6330-1971, Indian Standard - Recommendation for.
liquid flow measurement in open channels by weirs and
flumes - End depth method for estimation of flow in
rectangular channels with a free overfall (Approximate
method), December 1971.
10.22 IS: 9117-1979, Indian Standard - Recommendation for
liquid flow measurement in open channels by weirs and
f1umesEnd depth method for estimation of flow in non
rectangular channels with free overfall (Approximate
method).
10.23 IS: 9108-1979, Indian Standard-Liquid flow
measurements in open channels using thin plate weirs.

210
10.24 IS : 9116-1979, Indian Standard -Water Stage recorder
(Float type)
10.25 IS: 9163 (Part I) - 1979, Indian Standard - Dilution
methods for measurement of steady flow; Part I -
Constant rate injection method.
10.26 Hiranandani: M. G. and Chitale, S. V. 'Stream Gauging',
Second edition, Central Water and Power Research
Station, Pune, 1964.
10.27 World Meteorological Organisation, 'Manual of Stream
Gauging', Volume I 'Field Work'; Volume 1I,'Computation
of Discharge', Operational Hydrology Report No. 13,
WMO-No. 519, 1980.

211
 Chapter 11

MATHEMATICAL MODELLING
Rapid development is taking place in recent years in the
field of Mathematical modeling in the advanced hydraulics
laboratories in the world. The development is aimed at the
simulation of water and sediment flows; transport and dispersion
of pollutants in open channels and in wide shallow water bodies
using the digital computers.

The time varying hydraulic parameters in river and canals

such as velocity, water level, depth of flow, concentration of
sediment and pollutants, etc. and their time dependent changes,
changes in bed levels, bed material composition, changes in the
concentration of pollutants, etc are simulated under known
conditions and are then predicted in the changed conditions after
imposing the man made changes in the system.

Flexibility, accuracy and cost effectiveness play a vital

role in the development of the mathematical modeling techniques
and have helped to reduce dependence on physical models. In
fact, in the areas like flood routing, morphological modeling,
environmental modeling, etc use of mathematical model has now
become unavoidable.

Eminent mathematicians developed the basic

considerations more than two centuries before. Due to limitations
of the tools, numerical solutions based mainly on hand calculating
devices were then possible. Practical use of mathematical
models then was mainly restricted to analyse the data with the
help of statistical formulation and offer a solution in explicit
empirical forms.

The whole scenario has changed in the last 35-40 years

due to the development of computers. The mathematical modeling
techniques have now been widely accepted, which offer quick,
accurate and dependable solutions of hydraulic, morphological
or environmental phenomena. Now, the complexity and volume
of computations involved is not an important consideration.

212
 While adopting the mathematical modeling technique,
the original system of mathematical equations is simplified
keeping in view the desired objectives. The terms playing
insignificant contribution to the flow phenomena are neglected
and a satisfactory numerical formulation is made. The modified
formulation is made suitable for solution of a particular desired
flow phenomena. Such modifications are then cost effective and
are suitable for the speed and capacity of the computers.
However, the application would also get restricted to the desired
objectives kept in view by the designer.

Idea for presenting this chapter is to impart only the

preliminary ideas regarding the mathematical models. The
mathematical aspects like the original mathematical formulation,
simplifications made, adaptation of the equation to specific
objectives, etc are not discussed here. Moreover, the discussions
cover mainly the hydrodynamic phenomena, which is of more
interest to the bridge engineers. Other aspects, like evolving the
boundary conditions etc similar for both physical and mathematical
model studies. Therefore, the same also has not been covered
in the discussions.

11.1 TYPE OF MODELS

Depending up on the requirements of the studies and

the predominant hydraulic phenomena involved, mathematical
models can be grouped in to three types.

(a) Unsteady flow conditions

(b) Quasi-steady flow conditions
(c) Steady flow conditions

Normally, steady flow conditions are considered for

backwater computations, which are meant for a specific discharge
stage. For morphological computations, i.e. to compute time
dependent changes in the bed levels, quasi-steady flows are
used and computations are made at every discharge stage.
Unsteady flows are required to compute flood wave propagations,
flows in tidal reaches, etc.

Depending upon the requirements of studies,

mathematical models are also grouped in three types

213
(i) One dimensional model
(ii) Two dimensional model
(iii) Three dimensional model

One-dimensional model generally means modeling of a

river along its length. All the hydraulic parameters are averaged
and are assumed to act at one point. This is good approximation
for, say, flood propagation in the rivers, where variation in the flow
direction is generally important. In such computations, the bed
friction governs the speed of propagation and depth of flow at a
particular location, which is introduced by Manning or Chezy's
relations.

Main advantage of 1D models is the reproduction of a

network of channels in a river system. For example, Fig 11.1
shows a typical network of river channels reproduced in 1 D
mathematical model. Propagation of floods can be easily
reproduced in such complicated layout of river channels.

Natural changes like land slides, earth quakes, extreme

floods, deforestation or man made changes like construction of

Fig 11.1 : Typical network of river channels

reproduced in 1 D mathematical model

214
dams, barrages, bridges, embankments affect the natural regime
of rivers. Changes in the river regime affect river plan form,
hydraulic relations at a particular location etc. Modeling of
morphological processes in rivers is covered in this type. Using
quasi-steady flow conditions and sediment transport relations,
time dependent changes in the bed levels can be computed in 1
dimensional model. Fig 11.2 shows bed level changes computed
for an under-nourished canal for a period of about 30 years.
Progressive reduction in the overall bed levels and changes in
the bed profiles are apparent from the figure.

Fig 11.2 : Computed bed level profile using 1 D

morphological model

Two-dimensional model can be of two types, laid in X-Y

plane and in X-Z plane. Modeling in horizontal direction (X-Y plane)
computes variation in X-Y direction with vertically averaged values

Fig 11.3 : Bed levels (topography) reproduced

in a 2 D mathematical model.
215
of velocities at each location in X-Y plane. For bridge engineers,
it is necessary to assess the effect of the location / orientation of
a bridge, and the effect of constriction on both upstream and
downstream of a bridge. As far as these aspects are concerned,
the studies with the help of physical models could be comparable
with those conducted using 2D mathematical models.

For many bridge engineering solutions, the results of 2

D mathematical models are sufficient for the design of hydraulic
structures. Being in 2 D (plan) shape, good graphics help the
user to get more "natural and realistic" feel of the problem /
results. Few typical cases are presented here for quick
understanding. Fig 11.3 shows bed levels (topography) reproduced
in a 2 D mathematical model. Fig 11.4 shows velocities computed
in 2 D model for a specific discharge stage for the same
topography shown in Fig 11.3. The concentration of flow and
higher velocities generated at the nose of a spur result in higher
sediment transport and deeper scours immediately downstream
of the nose of a spur. Fig 11.5 shows bed levels computed at the
nose of a spur in sediment transport module of 2 D mathematical
model. The scours computed in the model have been reversed
for better understanding of the scours. These are some of the
examples of the use of mathematical models for the studies in
river engineering problems.

Fig 11.4 : Velocities computed in 2 D model for a specific

discharge stage

216
 Fig 11.5 : Bed levels computed at the nose of a spur in
sediment transport module of 2 D mathematical model.
The scours computed in the model have been reversed
for better understanding of the scours.

Modeling in vertical direction computes variation in X-Z

direction, whereas, variation in Y direction is not considered.
These models can be compared with the flume studies conducted
as a part of physical model studies. For bridge engineers, these
models may have use in specific cases like effect of bridge piers
on another bridge located in the vicinity, design of complex type
of piers, bridge laid at an angle to the main flow direction, etc.

The three-dimensional model considers X-Y-Z directions

simultaneously. Scours around bridge piers and abutments are
the typical problems 3 D in nature. For bridge engineers, the use
of 3 D mathematical models could be limited in the present
context.

Apart from the hydrodynamic modeling of river flows,

river engineering softwares are now commercially developed to
cover many other aspects like rainfall-runoff and other hydrological
aspects, ground water flows, dispersion of pollutants in the flows,
etc to bring other related aspects under "one umbrella". With the
development of graphics and application of many analytical
processes, graphical and pictorial presentation of the data,
analysis and display of the computed results, animation of results
at specific locations over specific period, etc has become highly
interesting and easy.
217
11.2 SYSTEMATIC STAGES FOR THE STUDIES

The data requirement for a physical model and a

mathematical model is exactly same. The development of a
physical or a mathematical model undergoes the same stages
as indicated below :

a. Layout of a physical system of desired reach.

b. Determination of boundary conditions at both upstream
and downstream ends.
c. Determination of initial conditions.
d. Initialization of model with pre-determined initial
conditions.
e. Accurate control of boundary conditions during the cali
bration run / run under existing conditions.
f. Check for smooth running of the model during the whole
computational domain.
g. Extraction of desired parameters during the model run.
h. Verification of results and comparison with prototype data.
i. Running the model under modified conditions as required
for the prediction run and extraction of desired
parameters.
j. Analysis of computed results, comparison with the
results under existing conditions and evaluation of the
results with modified conditions.

For transformation of the physical system into a

mathematical model, the discretisation technique is followed.
For one-dimensional mathematical model, the river length is
broken into fragments. Storage areas, flood planes etc are also
reproduced as per the requirement of the model. This part of
modeling technique yet required experience and skill. Water levels
and discharges are taken as "open boundaries" and suitably
specified as gauge discharge curves, water level hydrographs,
discharge hydrographs, etc.

11.3 ADVANTAGES AND DISADVANTAGES

A comparison between the two techniques is inevitable

which would also help to select the suitable one for a particular
problem posed.

218
a. Scale effects

Advantage of similarity of predominant forces is taken in

a physical model. Thus, the remaining forces are not similar and
bear scale effects. Interpretation of results needs proper
interpretation and evaluation of the values of different parameters
to the prototype. This has a particular reference to the sediment
transport, deposition, and scour etc. Interpretation of these
phenomena is more of an art than a standard technique. In a
mathematical model there are no scale effects as the natural
dimensions are used everywhere.

b. Type of models utilised

The physical models are used mainly for reproduction of

hydraulic phenomena, which are very well described. However,
reproduction of unsteady flow conditions in physical models is
difficult and costly. In case of channel network system, it is even
impracticable. Due to complexity of nature of sediment transport,
achieving similarity of sediment transport in a physical model is
not possible yet. Therefore, reproduction of year-to-year changes
in a physical model is not possible with the present techniques.
Mathematical models have some advantage over physical models
in such cases.

c. Cost and time

Layout of both physical model and mathematical model

is a matter of art. However, the time spent and cost involved in
analysis of data and adoption of the data for layout of the model
is much less in case of mathematical model.

Conducting model studies is both costly and time

consuming in case of physical models compared to the
mathematical models.

d. Establishment cost

Though the computers and their essential peripheral are

costly for initial installations, the same computers are used for
any mathematical model later on. Against this, costly ground
spaces and other supporting devices are required for physical

219
model which increase the preliminary cost before taking up the
construction of a physical model.

e. Model limitations

Description of 3 D flow phenomena is difficult and costly

with the present techniques in a mathematical model. A physical
model is a better alternative at present. Description of a
morphological process of a river is possible in a mathematical
model. A physical model is unable to reproduce long-term changes
in the riverbed. Therefore, a combination of both mathematical
and physical model is advisable for more complicated problems
and problems at the delicate and sensitive areas where support
of another technique is needed before finalising a decision by
one technique. In other cases, any suitable alternative could be
chosen for solution to a problem.

11.4 CLOSING REMARKS

There are many mathematical models available in the

market covering one or more of the discussed types. Keeping in
mind the advantages and limitations of the available models, they
can be conveniently used for studies.

220
REFERENCES

11.1 Cunge J.A., Holey F.M. - Practical aspects of

computational river hydraulics, Pitman Publishing Ltd.,
London - 1980

11.2 Verway A. - Lecture notes on mathematical modeling.

International Institute for Hydraulic and Environmental
Engineering, Delft, The Netherlands - 1984

11.3 Amein M. Feng C.S. - Implicit flood routing in natural

channels. Journal of Hydraulics Division, ASCE, Vol. 96,
HY 12, Proceedings, Paper 7773-1970

11.4 De. Vries M. - Solving river problems by hydraulic and

mathematical models. Publication No. 76, Delft
Hydraulics Laboratory - 1971

11.5 Thakar V.S. - Mathematical modeling of flows and water

quality phenomena. Diamond Jubilee Symposium of
CWPRS, Vol I, 1977

221
 LIST OF FIGURES & PHOTOS

Photo 1.1 : River Ravi and River Pubbar in the hilly

region of Himachal Pradesh
Photo 1.2 : Remnants of a suspension bridge lost during
the catastrophe
Photo 1.3 : Tributary of River Sutlej in Himachal Pradesh
Photo 2.1 (A) : Showing Brahmaputra River at Guwahati.
Straight course of the river is due to hillocks
located on both banks. Showing straight
chhanel of Kalang River, Assam
Photo 2.1 (B) : Examples of river meanders
Photo 2.1 (C) : Examples of complex type of river channel
Fig. 2.2 : Meander shapes developed using Circular
Arcs
Fig. 2.3 : Definition Sketch For Meander Dimensions
Photo 2.4 : Showing development of cut-off observed in
Gango River, Assam and formation of
Oxbow lakes
Fig. 2.5 : Cycle Changes in Ganga River at Mansi
Fig. 2.6 : Progressive Erosion in Flat And Acute
Bends
Fig. 2.7 : Superimposed courses of River Ganga at
Patna, Bihar showing nodal points
Photo 2.8 : Examples of braided river channels
Photo 2.9 : Progressive development of channels,
shifting of islands and consequent bank
erosion in river Brahmaputra, Assam
Fig. 2.10 : Shifting Courses of Kosi River
Fig. 2.11 : Training of Danube River in Hungary
Fig. 2.12 : Schematic Depiction of some of the types
of bed forms
Fig. 2.13 : Relation of Stream Power and Median Grain
Size to form of Bed Roughness

222
Fig. 3.1 : Two types of orbits followed by Indian
satellites
Fig. 3.2 : Method of observation by satellite
Photo 3.3 : The synoptic coverage. The right side
photographs are the satellite imageries
covering about 150 km (upper) and about 10
km (lower) of river length. Index maps are
on the left.
Photo 3.4 : The repetitive coverage. Imageries taken at
an interval of four years show the river
channel changes.
Photo 3.5 : Showing inaccessible river reach and
corresponding satellite image.
Photo 3.6 : Showing imageries of low and higher
resolution covering the same area.
Photo 3.7 : Showing the river reach observed by different
bands
Fig. 4.1 : Vertical velocity distribution with different
bottom roughnesses
Fig. 4.2 : Discharge computation for a river with
composite cross section
Fig. 4.3 : Relationship between specific energy, Depth
of Flow and Discharge Intensity
Fig. 5.1 : Railway Bridge on Chenab River at Sher
Shah
Fig. 5.2 : Guide Bunds of different shapes
Fig. 5.3 : Velocity distribution in channels of different
curvatures
Fig. 5.4 : Fitting at worst bend at curved head of a
guide bund
Fig. 5.5 : Composite curve for guide bund head of
Mokama Bridge on Ganga River
Fig. 5.6 : Thickness of pitching recommended by
Sethi
Fig. 5.7 : Size of Pitching Stone Vs Velocity
Fig. 5.8 : Size of Apron Stone Vs Velocity

223
Fig. 5.9 : Apron for Curved Head of a Guide Bund
with allowance for fanning out
Fig. 5.10 (A) : Apron as laid according to spring
Fig. 5.10 (B) : Apron as laid according to gales
Fig. 6.1 : Elliptical shape of guide bunds
Fig. 6.2 : Settling velocity of sediment grains of
different sizes
Fig. 6.3 : Apron design features
Fig. 6.4 : Design of Impermeable Spur
Fig. 6.5 : Depth of Bridge Piers according to Spring
Fig. 6.6 : Relation between Pier size and Scour Depth
given by Laursen
Fig. 7.1 : Elevation of Typical Porcupine
Photo 7.1 (A) : Typical Bamboo Porcupine
Fig. 7.2 : Sketch of Typical Crib
Photo 7.2 (A) : Bamboo crib constructed at site
Photo 7.3 : Bamboo structure projecting from the bank
into the river
Fig. 7.4 : Sketch of a typical permeable tree structure
Fig. 7.5 : Sketch of typical layout of Porcupine Spur
Fig 7.6 : Porcupine screens for inducing
sedimentation
Fig 7.7 (A) : Sketch of typical RCC porcupine
Photo 7.7(B) : RCC Porcupine laid at site
Photo 7.7(C) : Porcupine screen laid across a channel
Photo 8.1 : Internal structure of typical non-woven and
woven type geo-fabric
Photo 8.2 (A) : Showing geo-fabric being laid below the
slope pitching
Photo 8.2 (B) : Geo-fabric filter being laid below the stone
pitching. Cushion of sand is laid before
spreading the geo-fabric filter.
Photo 8.2 (C) : Geo-fabric filter being laid below the stone
pitching. Cushion of sand is laid on the top
of filter.
224
Fig. 10.1 : Staff-in-Pipe type Crest-Stage Gauge
Fig. 10.1 (A) : Crest Gauge
Fig. 10.2 : Pivot Point Layout
Fig. 10.3 : Air Line and Wet Line Corrections
Fig 11.1 : Typical network of river channels reproduced
in 1 D mathematical model
Fig 11.2 : Computed bed level profile using 1 D
morphological model
Fig 11.3 : Bed levels (topography) reproduced in a
2 D mathematical model.
Fig 11.4 : Velocities computed in 2 D model for a
specific discharge stage
Fig 11.5 : Bed levels computed at the nose of a spur
in sediment transport module of 2 D
mathematical model.

225
 LIST OF TABLES

Page No.
Table 2.1 : Meander relationships 18
Table 2.2 : Characteristics of Bed Forms in the 29
lower Reach of Brahmaputra River
Table 2.3 : River Formulae 32
Table 4.1 : Minimum clearances for bridges with 60
rectangular openings
Table 4.2 : Clearances for arch bridges 61
Table 4.3 : Calculation of scour depth 62
Table 4.4 : Comparison of actual widths of rivers 63
with those given by the Lacey Formula
Table 5.1 : Length of Guide Bund 72
Table 5.2 : Radius of Curved Head 75
According to Spring
Table 5.3 : Radius of Curvature and Angle of Sweep 77
According to Spring, Gales and Sethi
Table 5.4 : Radius of Upstream Curved Head of Guide 78
Bund According to Various Investigators
Table 5.5 : Thickness of Stone Protection 79
recommended by Spring
Table 5.6 : Thickness of Stone and Soiling for Slope 80
for Guide Bunds According to Gales
Table 5.7 : Thickness of stone pitching for shank 82
according to Sethi
Table 5.8 : Relationships for Size and Weight of Stone 85
Table 5.9 : Estimation of Depth of Scour 91
according to Gales
Table 5.10 : Estimation of Depth of Scour according to 91
committee headed by Khosla
Table 5.11 : Estimation of Depth of Scour according to 92
Sethi
Table 5.12 : Thickness of Launched Apron according to 93
Spring

226
Table 5.13 : Thickness of Launched Apron according to 94
Gales
Table 6.1 : Foundation Design Discharge with 125
Percentage Addition to Waterway
Design Discharge
Table 6.2 : Thickness of Pitching 138
Table 6.3 : Scour Factors for Apron Design in case 139
of Spurs
(A) According to Indian Standard Specifications
(B) On Basis of Experiments conducted by
Mushtaq Ahmed
Table 6.4 : Formulae in Vogue in the U.S.A. for 140
Prediction of Scour at the Noses of Spurs
Table 'A' : Scour level computations for bed material 151
sizes given in Alternative (A)
Table 6.5 : Etcheverry’s Maximum Allowable 155
Tractive Forces Given by Lane
Table 6.6 : USBR Limiting Tractive Forces in kg/cm2 155
Table 10.1 : Air Line Correction and Wet Line Correction 197
Table 10.2 : Values of Rugosity Coefficient n for 203
open channels with Relatively Coarse
Bed Material not characterized by Bed Form
Table 10.3 : Values of Rugosity Coefficient n for 203
open channels with other than Coarse
Bed Material

227
 INDEX

Page No.
Afflux
In Alluvial rivers 58
In rivers with non-scourable bed 55
Due to obstruction 58

Bridges
Alignment 66
Lacey Formulae 61
Location 66
Waterway 67, 112

Clearance 60

Discharge measurement 53, 192

Acoustic Method 208
Discharge for waterway 125
Discharge for foundation 124
Dilution method 207
Electromagnetic method 208
Mid section method 200
Moving Boat method 207
Radio tracer technique 208
Slope area method 123
Stage-discharge method 123
Velocity-area method 192, 202, 207

Filters
Graded 175
Geotextile 175, 178
Construction method 179

Free Board 59
Modelling
Couchy Number 183
Data requirement 158
Dimensional analysis 185
Dynamic similarity 182, 183

228
 Froude Number 28, 30, 140, 183
Geometric similarity 182
Kinetic similarity 182
Hydraulic modeling 182
Mathematic modeling 212
Mobile bed geometrically similar models 184
Mobile bed vertically exaggerated models 184
Model Limitations 185
Model scale designs 183
Modelling technique 182
Pie (p) theorem 185
Reynolds Number 183
Rigid bed geometrically similar models 184
Rigid bed vertically exaggerated models 184
Similitude requirements 182
Weber Number 183

Protection and Training Works 122

Data required for designs 158
Principles of - 52

Types of Protection & Training Works 66

Approach Banks 112, 141
Revetments 115, 145
Size & thickness of pitching stone 116
Types of revetments 115

Artificial cut-offs 117, 146

Bridge Pier protection 117
Shallow piers 117, 156
Scour at piers 62
Grip length for piers 147

Closure Bunds 117

Guide Bunds 63, 127

Apron as laid 97, 98, 135
Apron protection 90, 133
Apron quantity 93, 134
Construction programme 100

229
 Cross-section 101, 102, 129
Construction material 130
Curved head 74, 129
Length 71, 127
Scour along guide bund 84, 125
Shape in plan 69, 70, 127
Side slope and its protection 79, 129
Size of pitching stone 130
Thickness of pitching stone 132
Top level 129
Top width 129

Marginal Embankments 112, 144

Spurs and Groynes 100, 135

Cross-section 101, 102, 137
Function 100
Height 105
Inclination to bank line 105
Length 105
Location 104, 135
Material of construction 102
Permeable spurs 163
Porcupines & Crib 164, 165, 170
Porcupines (RCC) 173
Spacing of spurs 105
Types of spurs 100, 135

Stone size to withstand velocity 85, 88

Submontane region protection works 118

Scour in Rivers
Effect of bed forms 154
Flood scour 153
Lacey Formulae 61
Measurement of
Air line correction 197
Echo sounders 199
Various methods 200

230
 Wet line correction 197
Scour in alluvial rivers 58
Scour due to constriction 154
Scour at piers in clayey strata 154
Scour depth with different bed material at various
depths 147

Remote Sensing
Process 38
Characteristics 42
Data acquisition 44, 47

Rivers
Bed forms 24
Antidunes 27
Bed forms in Brahmaputra 29
Bed forms in Mississippi 28
Dunes 27
Flat beds 28
Implications of bed forms 28
Ripples 27
Transitions in bed forms 30

Behaviour of Rivers 2
Braiding pattern 22
Cut offs 13
Meanders 11
Meander movements 13
Meander implications 20
Meander relationships 18
River Formuale 30

River Reaches 1

Flood plains 5
Submontane 4
Tidal 6
Upper 1
River Types 1
Aggrading 6

231
 Braided 22
Degrading 6
Flashy 7
Himalayan 9
Meandering 11
South Indian 9
Stable 5
Virgin 8

River Water Level (Gauge or Stage)

Site requirement 190

Types of Gauges 190

Bubble 191
Chain 191
Crest 192
Float Well 191
Pneumatic 191
Staff 190
Stilling well 190
Tape 191
Wire 191

Velocity Measurements
Current meters 199
Distribution over a vertical 53
Formulae 30, 53
Mean velocity 200
Reduction coefficient 200
Rugosity coefficient in Manning formula 203

232
For suggestions, please write to:
mail@iricen.gov.in
 Published by :
INDIAN RAILWAYS INSTITUTE OF CIVIL ENGINEERING
11-A, South Main Road, Koregaon Park, Pune - 411001.

90/-